Parallels Server Bare Metal
5.0
User's Guide
Copyright © 1999-2011 Parallels Holdings, Ltd. and its affiliates. All rights reserved.
Parallels Holdings, Ltd.
c/o Parallels International GMbH.
Parallels International GmbH
Vordergasse 49
CH8200 Schaffhausen
Switzerland
Tel: + 41 526320 411
Fax: + 41 52672 2010
www.parallels.com
Copyright © 1999-2011 Parallels Holdings, Ltd. and its affiliates. All rights reserved.
This product is protected by United States and international copyright laws. The product’s underlying technology,
patents, and trademarks are listed at http://www.parallels.com/trademarks.
Microsoft, Windows, Windows Server, Windows NT, Windows Vista, and MS-DOS are registered trademarks of Microsoft
Corporation.
Apple, Mac, the Mac logo, Mac OS, iPad, iPhone, iPod touch, FaceTime HD camera and iSight are trademarks of Apple
Inc., registered in the US and other countries.
Linux is a registered trademark of Linus Torvalds.
All other marks and names mentioned herein may be trademarks of their respective owners.
Contents
Introduction ............................................................................................................... 9
About This Guide ............................................................................................................ 9
Organization of This Guide .................................................................................................... 10
Documentation Conventions ................................................................................................. 11
Getting Help.................................................................................................................. 12
Feedback...................................................................................................................... 13
Learning Parallels Server Bare Metal 5.0 Basics ................................................... 14
Parallels Server Bare Metal 5.0 Overview ....................................................................... 15
OS Virtualization Layer................................................................................................... 16
Basics of OS Virtualization ..................................................................................................... 17
Parallels Containers ............................................................................................................... 17
Virtuozzo File System ............................................................................................................ 18
Templates ............................................................................................................................. 18
Parallels Server Bare Metal Configuration .............................................................................. 19
Hardware Virtualization Layer......................................................................................... 19
Basics of Hardware Virtualization ........................................................................................... 20
Parallels Virtual Machines ...................................................................................................... 21
Virtual Machine Hardware ...................................................................................................... 22
Virtual Machine Files .............................................................................................................. 23
Support of Virtual and Real Media ......................................................................................... 24
Parallels Management Console ............................................................................................. 25
Resource Management ................................................................................................. 26
Understanding Licensing ............................................................................................... 26
Physical Server Availability Considerations ..................................................................... 27
Managing Virtual Machines and Containers ........................................................... 28
Creating a Virtual Machine and Container....................................................................... 29
Supported Guest Operating Systems .................................................................................... 31
Compatibility With Parallels Products..................................................................................... 32
Choosing a Container ID ....................................................................................................... 33
Choosing OS EZ Template .................................................................................................... 34
Contents
Performing Initial Configuration ...................................................................................... 34
Configuring Network Settings ................................................................................................ 35
Setting Passwords for Virtual Machines and Containers ........................................................ 36
Setting Startup Parameters ................................................................................................... 36
Starting, Stopping, and Querying Status of a Virtual Machine and Container ................... 37
Listing Virtual Machines and Containers ......................................................................... 38
Storing Extended Information on a Virtual Machine and Container .................................. 39
Copying a Virtual Machine and Container Within the Server ............................................ 40
Suspending a Virtual Machine and Container ................................................................. 42
Running Commands in a Virtual Machine and Container................................................. 43
Deleting a Virtual Machine and Container ....................................................................... 44
Managing Virtual Machine and Container Backups ......................................................... 44
Backups Overview ................................................................................................................. 45
Using pctl backup and pctl restore ........................................................................................ 47
Using pbackup and prestore ................................................................................................. 50
Migrating Virtual Machines and Containers..................................................................... 54
General Migration Requirements ........................................................................................... 55
Migrating Virtual Machines and Containers Between Parallels Servers ................................... 56
Migrating Containers to Virtual Machines ............................................................................... 61
Migrating Physical Computers to Virtual Machines and Containers ........................................ 64
Migrating Virtual Machines to Containers ............................................................................... 70
Migrating Xen Virtual Machines .............................................................................................. 71
Performing Container-Specific Operations ..................................................................... 72
Setting Name for Container ................................................................................................... 73
Moving Container Within the Parallels Server ......................................................................... 74
Disabling Container ............................................................................................................... 75
Reinstalling Container ............................................................................................................ 76
Performing Virtual Machine-Specific Operations ............................................................. 80
Viewing Detailed Information About Virtual Machines ............................................................. 80
Pausing a Virtual Machine...................................................................................................... 81
Managing Snapshots............................................................................................................. 82
Managing Templates ............................................................................................................. 86
Managing Virtual Machine Disks ............................................................................................ 88
Managing Virtual Machine Devices ........................................................................................ 93
Contents
Making Screenshots ............................................................................................................ 101
Assigning USB Devices to Virtual Machines ......................................................................... 102
Mounting Virtual Machines................................................................................................... 104
Configuring IP Address Ranges for Host-Only Networks ..................................................... 105
Converting Third-Party Virtual Machines and Disks.............................................................. 106
Enabling VNC Access to Virtual Machines ........................................................................... 108
Managing Resources ............................................................................................ 109
What are Resource Control Parameters? ..................................................................... 109
Managing CPU Resources .......................................................................................... 110
Managing CPU Shares and Limits ....................................................................................... 111
Configuring the Number of CPUs ........................................................................................ 113
Configuring CPU Affinity for Virtual Machines and Containers .............................................. 115
Managing Disk Quotas ................................................................................................ 115
What are Disk Quotas?........................................................................................................ 116
Disk Quota Parameters ....................................................................................................... 116
Turning On and Off Per-Container Disk Quotas ................................................................... 117
Setting Up Per-Container Disk Quota Parameters ............................................................... 118
Turning On and Off Second-Level Quotas for a Container ................................................... 119
Setting Up Second-Level Disk Quota Parameters ............................................................... 120
Checking Quota Status ....................................................................................................... 121
Cleaning Up Containers ....................................................................................................... 122
Managing Network Accounting and Bandwidth ............................................................ 125
Network Traffic Parameters ................................................................................................. 126
Configuring Network Classes .............................................................................................. 127
Viewing Network Traffic Statistics ........................................................................................ 129
Turning On and Off Network Bandwidth Management ........................................................ 131
Configuring Network Bandwidth Management .................................................................... 133
Managing Disk I/O Parameters .................................................................................... 134
Configuring Priority Levels for Virtual Machines and Containers ........................................... 135
Configuring the Disk I/O Bandwidth ..................................................................................... 136
Viewing Disk I/O Statistics ................................................................................................... 138
Managing Memory Parameters for Containers.............................................................. 139
Configuring Main VSwap Parameters .................................................................................. 140
Configuring the Memory Allocation Limit .............................................................................. 141
Tuning VSwap ..................................................................................................................... 142
Contents
Configuring Legacy Containers ............................................................................................ 143
Managing Memory Resources for Virtual Machines ...................................................... 144
Configuring Main Memory Resources .................................................................................. 144
Configuring Additional Memory Parameters ......................................................................... 145
Enabling Memory Hotplug for Virtual Machines .................................................................... 148
Managing Container Resources Configuration.............................................................. 149
Splitting Server Into Equal Pieces ........................................................................................ 150
Scaling Container Configuration .......................................................................................... 151
Validating Container Configuration ....................................................................................... 152
Applying New Configuration Sample to Container ............................................................... 153
Managing Services and Processes ...................................................................... 154
What Are Services and Processes ............................................................................... 155
Main Operations on Services and Processes................................................................ 156
Managing Processes and Services .............................................................................. 157
Viewing Active Processes and Services ............................................................................... 158
Monitoring Processes in Real Time ...................................................................................... 160
Changing Services Mode..................................................................................................... 161
Determining Container Identifiers by Process IDs ................................................................ 162
Starting, Stopping, and Restarting Services ........................................................................ 162
Managing Parallels Server Bare Metal Network ................................................... 163
Managing Network Adapters on the Parallels Server .................................................... 164
Listing Adapters .................................................................................................................. 165
Creating VLAN Adapters ..................................................................................................... 166
Networking Modes in Parallels Server Bare Metal ......................................................... 166
Host-Routed Mode.............................................................................................................. 167
Virtual Network Mode .......................................................................................................... 170
Differences Between Host-Routed and Virtual Network Modes ........................................... 172
Configuring Virtual Machines and Containers in Host-Routed Mode.............................. 173
Configuring Virtual Machines and Containers in Virtual Network Mode .......................... 174
Managing Virtual Networks .................................................................................................. 174
Managing Adapters in Containers ........................................................................................ 179
Managing Adapters in Virtual Machines ............................................................................... 183
Managing Private Networks ......................................................................................... 187
Learning Private Networks ................................................................................................... 187
Contents
Setting Up Private Networks ................................................................................................ 191
Managing Licenses ............................................................................................... 193
Installing the License ................................................................................................... 194
Updating the Current License ...................................................................................... 195
Transferring the License to Another Server ................................................................... 196
Viewing the Current License ........................................................................................ 197
Viewing the License ............................................................................................................. 198
License Statuses ................................................................................................................. 200
Keeping Your System Up To Date ........................................................................ 201
Updating Parallels Server Bare Metal Software ............................................................. 202
Updating in Graphical Mode ................................................................................................ 203
Updating in Command-Line Mode ...................................................................................... 209
Updating Software In Virtual Machines ......................................................................... 210
Updating Containers ................................................................................................... 210
Updating EZ Template Packages Inside a Container ........................................................... 211
Updating OS EZ Template Caches ...................................................................................... 212
Advanced Tasks .................................................................................................... 213
Configuring Capabilities ............................................................................................... 213
Creating VZFS Symlinks Inside a Container ......................................................................... 214
Available Capabilities for Containers .................................................................................... 215
Creating Customized Containers ................................................................................. 218
Using Customized OS EZ Templates ................................................................................... 219
Using EZ OS Template Sets ................................................................................................ 221
Using Customized Application Template ............................................................................. 223
Changing System Time From Containers ..................................................................... 225
Obtaining Server ID From Inside a Container ................................................................ 226
Enabling VPN for Containers........................................................................................ 226
Managing Server Resources Parameters ..................................................................... 227
Setting Immutable and Append Flags for Container Files and Directories ...................... 228
Customizing the /proc/meminfo Output in Containers .................................................. 229
Loading iptables Modules ............................................................................................ 231
Loading iptables Modules to Parallels Server ....................................................................... 231
Loading iptables Modules to Containers .............................................................................. 232
Contents
Creating Configuration Files for New Linux Distributions ............................................... 234
Monitoring Resources ................................................................................................. 235
Aligning Disks and Partitions in Virtual Machines .......................................................... 236
Running Parallels Server Bare Metal 5.0 in Virtual Machines ......................................... 242
Troubleshooting .................................................................................................... 244
General Considerations ............................................................................................... 245
Kernel Troubleshooting................................................................................................ 247
Using ALT+SYSRQ Keyboard Sequences ........................................................................... 247
Saving Kernel Faults (OOPS) ............................................................................................... 248
Finding a Kernel Function That Caused the D Process State ............................................... 249
Problems With Container Management ........................................................................ 249
Failure to Start a Container .................................................................................................. 250
Failure to Access a Container From Network ....................................................................... 251
Failure to Log In to a Container............................................................................................ 251
Getting Technical Support ........................................................................................... 252
Preparing and Sending Questions to Technical Support ...................................................... 252
Submitting Problem Report to Technical Support ................................................................ 253
Establishing Secure Channel to Parallels Support ................................................................ 255
Glossary ................................................................................................................. 257
Index ...................................................................................................................... 259
CHAPTER 1
Introduction
This chapter provides basic information about Parallels Server Bare Metal 5.0 and this guide.
In This Chapter
About This Guide ................................................................................................... 9
Getting Help ........................................................................................................... 12
Feedback ............................................................................................................... 13
About This Guide
The Parallels Server Bare Metal 5.0 User's Guide provides comprehensive information on Parallels
Server Bare Metal 5.0 - high-end virtualization software for bare metal servers. It covers the
necessary theoretical conceptions as well as practical aspects of working with Parallels Server Bare
Metal. The guide will familiarize you with the way to create and administer virtual machines and
Containers using the Parallels command line interface.
Note: The guide does not explain how to install and configure your Parallels Server Bare Metal 5.0
system. For detailed information on these operations, see the Parallels Server Bare Metal 5.0 Installation
Guide.
The primary audience for this guide is anyone responsible for administering one or more systems
running Parallels Server Bare Metal 5.0. We assume that you have some familiarity with how to
work in the Linux command line.
Introduction
Organization of This Guide
This guide is organized in the following way:
•
Chapter 1, Introduction, gives an overview of the Parallels Server Bare Metal product and this
guide.
•
Chapter 2, Parallels Server Bare Metal Basics, explains the general principles of Parallels
Server Bare Metal operation.
•
Chapter 3, Operations on virtual machines and Containers, covers those operations that
you can perform on a virtual machine and Container: creating and deleting virtual machines and
Containers, starting and stopping them, backing up and restoring, etc. You will also learn how
to perform different kinds of migration: migrate virtual machines and Containers between
Parallels servers, migrate a physical server to a virtual machine and Container, and migrate a
Container to a virtual machine.
•
Chapter 4, Managing Resources, focuses on configuring and monitoring the resource control
parameters for virtual machines and Containers. These parameters comprise disk quotas,
network accounting and shaping, CPU and system resources.
•
Chapter 5, Managing Services and Processes, familiarizes you with the operations you can
perform on processes and services in Parallels Server Bare Metal.
•
Chapter 6, Managing Parallels Server Bare Metal Network, familiarizes you with the Parallels
Server Bare Metal network structure and explains how to manage networks in Parallels Server
Bare Metal systems.
•
Chapter 7, Managing Licenses, provides detailed information on managing licenses in
Parallels Server Bare Metal.
•
Chapter 8, Keeping Your System Up To Date, informs you of the ways to keep all the
software components of a Parallels server up to date.
•
Chapter 9, Advanced Tasks, enumerates those tasks that are intended for advanced system
administrators who would like to obtain deeper knowledge about Parallels Server Bare Metal
capabilities.
•
Chapter 10, Troubleshooting, suggests ways to resolve common inconveniences should they
occur during your work with Parallels Server Bare Metal.
10
Introduction
Documentation Conventions
Before you start using this guide, it is important to understand the documentation conventions used
in it.
The table below presents the existing formatting conventions.
Formatting convention
Special Bold
Type of Information
Example
Items you must select, such as
Go to the Resources tab.
menu options, command buttons,
or items in a list.
Titles of chapters, sections, and
subsections.
Read the Basic Administration chapter.
Italics
Used to emphasize the
These are the so-called EZ templates.
importance of a point, to
introduce a term or to designate a To destroy a Container, type vzctl
command-line placeholder, which destroy ctid.
is to be replaced with a real name
or value.
Monospace
The names of commands, files,
and directories.
Use vzctl start to start a Container.
Preformatted
On-screen computer output in
your command-line sessions;
source code in XML, C++, or
other programming languages.
Saved parameters for Container
101
Monospace Bold
What you type, as contrasted with # rpm –V virtuozzo-release
on-screen computer output.
Key+Key
Key combinations for which the
user must press and hold down
one key and then press another.
Ctrl+P, Alt+F4
Besides the formatting conventions, you should also know about the document organization
convention applied to Parallels documents: chapters in all guides are divided into sections, which,
in their turn, are subdivided into subsections. For example, About This Guide is a section, and
Documentation Conventions is a subsection.
11
Introduction
Getting Help
In addition to this guide, there are a number of other resources available for Parallels Server Bare
Metal which can help you use the product more effectively. These resources include:
Manuals:
• Parallels Server Bare Metal 5.0 Installation Guide. This guide provides detailed information
on installing Parallels Server Bare Metal on your server, including the pre-requisites and the
stages you shall pass.
• Getting Started With Parallels Server Bare Metal 5.0. This guide provides basic information
on how to install Parallels Server Bare Metal on your server, create new Containers and
virtual machines, and perform main operations on them. As distinct from the Parallels Server
Bare Metal 5.0 Installation Guide, it does not contain detailed description of all the
operations needed to install and set Parallels Server Bare Metal to work (e.g. installing
Parallels Server Bare Metal in text mode).
• Parallels Server Bare Metal 5.0 Templates Management Guide. This guide is meant to
provide complete information on Parallels templates - an exclusive Parallels technology
allowing you to efficiently deploy standard Linux applications inside your Containers and to
greatly save the physical server resources (physical memory, disk space, etc.).
• Parallels Command Line Reference Guide. This guide is a complete reference on all Parallels
Server Bare Metal configuration files and command line utilities.
• Deploying Clusters in Parallels-Based Systems. This guide describes the process of creating
Parallels failover and GFS clusters using the Red Hat Cluster Suite (RHCS) software.
Help systems:
• Getting Started with Parallels Management Console. This help system provides information
on how to start working in Parallels Management Console. You will learn how to install this
application on your computer, connect to a server running Parallels Server Bare Metal, and
perform the basic operations on your virtual machines.
• Parallels Management Console User's Guide. This help system provides detailed information
on Parallels Management Console - a graphical user interface tool for managing physical
servers and their virtual machines.
12
Introduction
Feedback
If you spot a typo in this guide, or if you have an opinion about how to make this guide more
helpful, you can share your comments and suggestions with us by completing the Documentation
Feedback form on our website (http://www.parallels.com/en/support/usersdoc/).
13
CHAPTER 2
Learning Parallels Server Bare Metal 5.0
Basics
This chapter provides a brief description of Parallels Server Bare Metal 5.0, Parallels virtual
machines and Containers, their specifications and underlying technologies.
In This Chapter
Parallels Server Bare Metal 5.0 Overview ................................................................ 15
OS Virtualization Layer ............................................................................................ 16
Hardware Virtualization Layer .................................................................................. 19
Resource Management .......................................................................................... 26
Understanding Licensing ........................................................................................ 26
Physical Server Availability Considerations .............................................................. 27
Learning Parallels Server Bare Metal 5.0 Basics
Parallels Server Bare Metal 5.0 Overview
Parallels Server Bare Metal 5.0 provides you with the possibility to simultaneously run Parallels
virtual machines and Containers on the same server. Using this software, you can efficiently use
your server's hardware resources by sharing them among multiple virtual machines and Containers.
Graphically, a Mac with the Parallels Server Bare Metal software installed can be represented as
follows:
15
Learning Parallels Server Bare Metal 5.0 Basics
At the base resides server hardware. Next is the Parallels Server Bare Metal software which is
installed directly on the server hardware and does not need any operating system for its
functioning. Parallels Server Bare Metal includes two virtualization layers:
•
Hardware virtualization layer. This layer provides the necessary environment for creating and
managing Parallels virtual machines.
•
OS virtualization layer. This layer provides the necessary environment for creating and
managing Parallels Containers.
For more information on both layers, see OS Virtualization Layer (p. 16) and Hardware
Virtualization Layer (p. 19).
Effectively uniting both virtualization technologies, Parallels Server Bare Metal provides the best
value for cost conscious organizations enabling them to:
•
standardize server hardware platforms
•
effectively consolidate server resources
•
consolidate and support legacy OSs and applications
•
streamline server and application deployment, maintenance, and management
•
simplify software testing and development
•
optimize server and application availability
Parallels Server Bare Metal allows you to create virtual machines and Containers and Containers
and manage them using the same tools you would use on systems running Parallels Server 3.0 and
Parallels Virtuozzo Containers 4.0. These tools include:
•
Command line interface (CLI). This tool comprises a set of Parallels command line utilities and
can be used to manage virtual machines and Containers both locally and remotely.
•
Parallels Management Console. Parallels Management Console is a remote management tool
for Parallels Server Bare Metal with a graphical user interface. This tool can be used to manage
physical servers and Parallels virtual machines residing on them.
Note: In this version of Parallels Server Bare Metal, you cannot use Parallels Management Console to
create and manage Parallels Containers.
OS Virtualization Layer
This section provides detailed information on the OS virtualization layer, one of the two components
of Parallels Server Bare Metal, responsible for providing support for Parallels Containers.
16
Learning Parallels Server Bare Metal 5.0 Basics
Basics of OS Virtualization
The OS virtualization allows you to virtualize physical servers on the operating system (kernel) layer.
The diagram below shows the basic architecture of OS virtualization.
The OS virtualization layer ensures isolation and security of resources between different Containers.
The virtualization layer makes each Container appear as a standalone server. Finally, the Container
itself houses its own applications and workload. OS virtualization is streamlined for the best
performance, management, and efficiency. Its main advantages are the following:
•
Containers perform at levels consistent with native servers. Containers have no virtualized
hardware and use native hardware and software drivers making its performance unbeatable.
•
Each Container can seamlessly scale up to the resources of an entire physical server.
•
OS virtualization technology provides the highest density available from a virtualization solution.
You can create and run up to 100s of Containers on a standard production physical server.
•
Containers use a single OS, making it extremely simple to maintain and update across
Containers. Applications may also be deployed as a single instance.
Parallels Containers
From the point of view of applications and Container users, each Container is an independent
system. This independence is provided by the Parallels Server Bare Metal OS virtualization layer.
Note that only a negligible part of the CPU resources is spent on virtualization (around 1-2%). The
main features of the virtualization layer implemented in Parallels Server Bare Metal are the following:
•
A Container looks like a normal Linux system. It has standard startup scripts; software from
vendors can run inside Containers without any modifications or adjustment.
•
A user can change any configuration file and install additional software inside Containers.
•
Containers are fully isolated from each other (file system, processes, sysctl variables) and
Parallels virtual machines.
•
Containers share dynamic libraries, which greatly saves memory.
•
Processes belonging to a Container are scheduled for execution on all available CPUs.
Consequently, Containers are not bound to only one CPU and can use all available CPU power.
17
Learning Parallels Server Bare Metal 5.0 Basics
Virtuozzo File System
Virtuozzo File System (VZFS) is a file system that allows sharing common files among multiple
Containers without sacrificing flexibility. It is possible for Container users to modify, update, replace,
and delete shared files. When a user modifies a shared file, VZFS creates a private copy of the file
transparently for the user. Thus, the modifications do not affect the other users of the file. Main
benefits of VZFS are the following:
•
VZFS saves memory required for executables and libraries. A typical Container running a simple
web site might consume around 20–30 MB of RAM just for executable images. Sharing this
memory improves scalability and total system performance.
•
VZFS saves disk space. A typical Linux OS installation occupies several hundred megabytes of
disk space. Sharing the files allows you to save up to 90% of disk space.
•
VZFS does not require having different physical partitions for different Containers or creating a
special “file system in a file” setup for a Container. This significantly simplifies disk
administration.
•
Disk quota enables the administrator to limit disk resources available to a Container on the fly.
Disk quota for users and groups inside Containers is also supported.
Templates
A template (or a package set) in Parallels Server Bare Metal is a set of original application files
repackaged for mounting over Virtuozzo File System. Usually it is just a set of RPM packages for
Red Hat like systems. Parallels Server Bare Metal provides tools for creating templates, installing,
upgrading, adding them to and removing them from a Container. Using templates lets you:
•
share the RAM among similar applications running in different Containers to save hundreds of
megabytes of memory
•
share the files comprising a template among different Containers to save gigabytes of disk
space
•
deploy applications simultaneously in many Containers
•
use different versions of an application on different Containers (for example, perform an
upgrade only in certain Containers)
There are two types of templates: OS templates and application templates. An OS template is an
operating system and the standard set of applications to be found right after the installation.
Parallels Server Bare Metal uses OS templates to create new Containers with a preinstalled
operating system. An application template is a set of repackaged software packages optionally
accompanied with configuration scripts. Application templates are used to add extra software to
existing Containers. For example, you can create a Container on the basis of the redhat OS
template and add the MySQL application to it with the help of the mysql template.
For detailed information on Parallels templates, see the Parallels Server Bare Metal Templates
Management Guide.
18
Learning Parallels Server Bare Metal 5.0 Basics
Parallels Server Bare Metal Configuration
Parallels Server Bare Metal allows you to flexibly configure various settings for the physical server in
general as well as for each and every Container. Among these settings are disk and user quota,
network parameters, default file locations and configuration sample files, and others.
Parallels Server Bare Metal stores all OS virtualization-related configuration information in two types
of files: the global configuration file /etc/vz/vz.conf and Container configuration files
/etc/vz/conf/<CT_ID>.conf. The global configuration file defines global and default
parameters for Container operation, for example, logging settings, enabling and disabling disk
quota for Containers, the default configuration file and OS template on the basis of which a new
Container is created, and so on. On the other hand, a Container configuration file defines the
parameters for a given particular Container, such as disk quota and allocated resources limits, IP
address and hostname, and so on. If a parameter is configured both in the global configuration file
and in the Container configuration file, the Container configuration file takes precedence. For a list
of parameters constituting the global configuration file and the Container configuration files, refer
the Parallels Server Bare Metal 5.0 Command Line Reference Guide.
The configuration files are read when the Parallels Server Bare Metal software and/or Containers
are started. However, Parallels Server Bare Metal standard utilities (for example, pctl) allow you to
change many configuration settings on the fly, either without modifying the corresponding
configuration files or with their modification (if you want the changes to apply the next time the
Parallels Server Bare Metal software and/or Containers are started).
Hardware Virtualization Layer
This section familiarizes you with the second component of Parallels Server Bare Metal - the
hardware virtualization layer. This layer provides the necessary environment for creating and
managing Parallels virtual machines.
19
Learning Parallels Server Bare Metal 5.0 Basics
Basics of Hardware Virtualization
Parallels Server Bare Metal is based on the concept of hardware virtualization. Hardware
virtualization has a base layer—a hypervisor. This layer is loaded directly on the bare server and
acts as an intermediary between the server hardware and virtual machines. To allocate hardware
and resources to virtual machines, Parallels Server Bare Metal virtualizes all hardware on the server.
Once virtualized, hardware and resources can be easily assigned to virtual machines. Based on the
virtual hardware, a virtual machine runs its own complete copies of an operating system and
applications.
The following diagram shows the basic architecture of hardware virtualization.
Like OS virtualization, hardware virtualization also provides many benefits the main of which are
listed below:
•
Create multiple virtual machines with different operating systems on a single physical computer.
•
Manage several physical servers at a time using Parallels Management Console, an integrated
GUI-based multi-server and cross-platform management tool.
•
Run several guest operating systems and their applications simultaneously on a single physical
computer without rebooting.
•
Consolidate and virtualize the computing environment, reduce hardware costs, lower operating
expenses, and increase productivity.
•
Use open APIs and SDK to extend management integration with in-house and third-party
applications.
20
Learning Parallels Server Bare Metal 5.0 Basics
Parallels Virtual Machines
From the point of view of applications and virtual machine users, each virtual machine is an
independent system with an independent set of virtual hardware. This independence is provided by
the Parallels Server Bare Metal hardware virtualization layer. The main features of the virtualization
layer are the following:
•
A virtual machine looks like a normal computer. It has its own virtual hardware, and software
applications can run in virtual machines without any modifications or adjustment.
•
A user can easily change the virtual machine configuration (e.g. add a new virtual disk or
increase memory).
•
Virtual machines are fully isolated from each other (file system, processes, sysctl variables)
and Parallels Server Bare Metal.
•
Install any of the supported operating systems in the virtual machine. The guest operating
system and its applications are isolated inside a virtual machine and share physical hardware
resources with other virtual machines.
Intel and AMD Virtualization Technology Support
Parallels Server Bare Metal provides support for Intel and AMD virtualization technologies
comprising a set of processor enhancements and improving the work of virtualization solutions.
Utilizing these technologies, Parallels Server Bare Metal can offload some workload to the system
hardware, which results in the "near native" performance of guest operating systems.
21
Learning Parallels Server Bare Metal 5.0 Basics
Virtual Machine Hardware
A Parallels virtual machine works like a stand-alone computer with the following hardware:
CPU
Up to 16-core Intel/AMD CPU
Motherboard
Intel i965 chipset-based motherboard
RAM
Up to 64 GB of main memory
Video Adapter
VGA and SVGA with VESA 3.0 compatible video adapter
Video RAM
Up to 256 MB of video memory
Floppy Disk Drive
1.44 MB floppy disk drive mapped to an image file or to a
physical floppy drive
IDE Devices
Up to 4 IDE devices
•
Hard Disk
Hard disk drive mapped to an image file (up to 2 TB each)
•
CD/DVD-ROM
Drive
CD/DVD-ROM drive mapped to a physical drive or to an image
file
SCSI Devices
Up to 15 SCSI devices
•
Hard Disk
Hard disk drive mapped to an image file (up to 2 TB each)
•
Generic SCSI
Device
Generic SCSI device
Network Interfaces
Up to 16 network interfaces, including Ethernet virtual network
cards compatible with RTL8029
Serial (COM) Ports
Up to 4 serial (COM) ports mapped to a socket or to an output
file
Parallel (LPT) Ports
Up to 3 parallel (LPT) ports mapped to an output file, to a real
port, or to a printer
Sound Card
AC'97-compatible sound card, sound recording support
Keyboard
Generic PC keyboard
Mouse
PS/2 wheel mouse
22
Learning Parallels Server Bare Metal 5.0 Basics
Virtual Machine Files
A virtual machine has at least two files: a configuration file (PVS file) and a hard disk image file (HDD
file). It can also have additional files: a file for each additional virtual hard disk and output files for
virtual ports. By default, the virtual machines files are stored in the /var/parallels directory on
the Parallels server.
The list of files related to a virtual machine is given in the table below:
File Name
Description
.pvm
A bundle that contains the virtual machine files.
.pvs
A virtual machine configuration file. It defines the hardware and resources
configuration of the virtual machine. The configuration file is automatically
generated during the virtual machine creation.
.sav
A dump file created when you suspend the virtual machine. This file contains the
state of the virtual machine and its applications at the moment the suspend was
invoked.
.mem
A file containing the memory dump for the suspended virtual machine. For a
running virtual machine, it is a temporary virtual memory file.
.hdd
A file representing a virtual hard disk. When you create a virtual machine, you can
create it with a new virtual hard disk or use an existing one. A virtual machine can
have several hard disks.
.iso
An image file of a CD or DVD disc. Virtual machines treat ISO images as real
CD/DVD discs.
.txt
Output files for serial and parallel ports. The output .txt files are generated when
a serial or parallel port connected to an output file is added to the virtual machine
configuration.
23
Learning Parallels Server Bare Metal 5.0 Basics
Support of Virtual and Real Media
This section lists the types of disks that can be used by Parallels virtual machines and provides the
information about basic operations you can perform on these disks.
Supported Types of Hard Disks
Parallels virtual machines can use only virtual hard disks image files as their hard disks.
Virtual Hard Disks
The capacity of a virtual hard disk can be set from 100 MB to 2 TB.
Virtual hard disks can be of either plain or expanding format. When you create a virtual machine in
Express Windows or Typical mode (in the New Virtual Machine wizard), the disk is created in the
expanding format.
plain
A plain virtual hard disk image file has a fixed size. The size is determined when the disk is
created. Plain disks can be created with the help of New Virtual Machine wizard (the Custom
mode.)
expandin
g
An expanding virtual hard disk image file is small initially. Its size grows as you add
applications and data to the virtual hard disk in the guest OS.
Split disks
A virtual disk of either format can be a single-piece disk or a split disk. A split disk is cut into 2 GB
pieces and is stored as a single .hdd file.
CD/DVD Discs and Their Images
Parallels Server can access real CD/DVD discs and images of CD/DVD discs.
Parallels Server has no limitations on using multi-session CD/DVD discs. A virtual machine can play
back audio CDs without any limitations on copy-protected discs.
If your server has a recordable optical drive, you can use it to burn CD or DVD discs in a virtual
machine.
Parallels Server supports CD/DVD disc images in ISO, CUE, and CCD formats.
Floppy Disks and Floppy Disk Images
Parallels Server can use two types of floppy disks:
•
24
Real diskettes inserted into a real floppy disk drive that is connected to the virtual machine.
Learning Parallels Server Bare Metal 5.0 Basics
•
Floppy disk image files having the .fdd extension and connected to the virtual machine.
Parallels Server treats floppy disk images like real diskettes. Parallels Server supports floppy disk
image files that have the .fdd extension and are 1.44 MB in size.
With Parallels Server, you can also create an image of a blank floppy using the Floppy Disk pane of
the Virtual Machine Configuration dialog.
Note: Parallels Server cannot create images of real diskettes.
Parallels Management Console
Parallels Management Console is a remote tool with a graphical user interface (GUI) for managing
your physical servers with Parallels Server Bare Metal and virtual machines residing on them. This
tool uses a typical client-server architecture.
The client application with the graphical user interface is installed on a computer running one of the
supported Linux or Windows operating systems. Once the client application is up and running, it
can connect to the Parallels Server Bare Metal software on a physical server. The client application
can control multiple physical servers simultaneously (e.g. Physical Server #1 and Physical Server #2
as shown in the picture above). After the connection to the required physical server has been
established, you can start managing this server and its virtual machines using the intuitive and
comfortable GUI.
25
Learning Parallels Server Bare Metal 5.0 Basics
Resource Management
Parallels Server Bare Metal resource management controls the amount of resources available to
virtual machines and Containers. The controlled resources include such parameters as CPU power,
disk space, a set of memory-related parameters. Resource management allows you to:
•
effectively share available physical server resources among virtual machines and Containers
•
guarantee Quality-of-Service in accordance with a service level agreement (SLA)
•
provide performance and resource isolation and protect from denial-of-service attacks
•
simultaneously assign and control resources for a number of virtual machines and Containers
•
collect usage information for system health monitoring
Resource management is much more important for Parallels Server Bare Metal than for a
standalone server since server resource utilization in such a system is considerably higher than that
in a typical system.
Understanding Licensing
To start using the Parallels Server Bare Metal software, you need a special license - Parallels Server
Bare Metal license. You must install this license on your server after or when installing Parallels
Server Bare Metal on it. Every physical server hosting virtual machines and Containers must have
its own license. Licenses are issued by Parallels and define a number of parameters in respect of
your physical server. The main licensed parameters are listed below:
•
The number of CPUs which can be installed on the physical server. Keep in mind that each of
the Dual Core and Hyperthreading processors is regarded as one CPU.
•
The license expiration date. Any license can be time-limited or permanent.
Parallels Server Bare Metal licenses have a start date, and if they are time-limited, can also have
an expiration date specified in them. You must set up your system clock correctly; otherwise,
the license validation may fail.
•
The number of virtual machines and Containers that can simultaneously run on he physical
server.
•
The platform and architecture with which the Parallels Server Bare Metal software is
compatible.
26
Learning Parallels Server Bare Metal 5.0 Basics
Physical Server Availability Considerations
The availability of a physical server running Parallels Server Bare Metal is more critical than the
availability of a typical PC server. Since it runs multiple virtual machines and Containers providing a
number of critical services, physical server outage might be very costly. It can be as disastrous as
the simultaneous outage of a number of servers running critical services.
To increase physical server availability, we suggest that you follow the recommendations below:
•
Use a RAID storage for critical virtual machines and Containers. Do prefer hardware RAIDs, but
software mirroring RAIDs might suit too as a last resort.
•
Do not run any software on the server itself. Create special virtual machines and Containers
where you can host necessary services such as BIND, FTPD, HTTPD, and so on. On the server,
you need only the SSH daemon. Preferably, it should accept connections from a pre-defined
set of IP addresses only.
•
Do not create users on the server itself. You can create as many users as you need in any
virtual machine and Container. Remember: compromising the server means compromising all
virtual machines and Containers as well.
27
CHAPTER 3
Managing Virtual Machines and Containers
This chapter describes how to perform day-to-day operations on virtual machines and Containers.
Note: We assume that you have successfully installed, configured, and deployed your Parallels Server
Bare Metal system. If you have not, refer to the Parallels Server Bare Metal Installation Guide for detailed
information on these operations.
In This Chapter
Creating a Virtual Machine and Container ................................................................ 29
Performing Initial Configuration ................................................................................ 34
Starting, Stopping, and Querying Status of a Virtual Machine and Container............. 37
Listing Virtual Machines and Containers................................................................... 38
Storing Extended Information on a Virtual Machine and Container ............................ 39
Copying a Virtual Machine and Container Within the Server...................................... 40
Suspending a Virtual Machine and Container ........................................................... 42
Running Commands in a Virtual Machine and Container .......................................... 43
Deleting a Virtual Machine and Container................................................................. 44
Managing Virtual Machine and Container Backups .................................................. 44
Migrating Virtual Machines and Containers .............................................................. 54
Performing Container-Specific Operations ............................................................... 72
Performing Virtual Machine-Specific Operations....................................................... 79
Managing Virtual Machines and Containers
Creating a Virtual Machine and Container
This section explains how to create a new Parallels virtual machine and Container. The options you
should pass to this command differ depending on whether you want to create a virtual machine or
Container.
Creating a Container
To create a Container, you can use the pctl create command. This command requires the
following parameters:
Argument
Description
Container ID
A numeric ID associated with a Container (101, 403, and so on). The
Container ID should be an integer greater than 100 and unique for a
given Parallels server.
OS template name
The name of the OS template to base your Container on. Parallels Server
Bare Metal is shipped with a number of ready-to-use OS templates. To
find out the names of the available templates, use the vzpkg list -O
command.
For the list of operating systems you can run in your virtual machines
and Containers, see Supported Guest Operating Systems (p. 31).
Configuration file
The name of the sample configuration file that will be used for setting all
the Container resource control parameters. The sample configuration
files are residing in the /etc/vz/conf directory on the physical server
and have names with the following mask: ve-<configname>.confsample. The most commonly used sample is the ve-basic.confsample file. This sample file has resource control parameters suitable
for most Containers.
Thus, for example, you can create a new Container by executing the following command:
# pctl create 101 --ostemplate fedora-core-9-x86 -–config basic
Creating Container private area (fedora-core-9-x86)
...
Container private area was created
In this case Parallels Server Bare Metal will create a Container with ID 101, the Fedora 9 OS
installed inside, and the configuration parameters taken from the ve-basic.conf-sample
sample configuration file.
Note: For more information on options you can pass to pctl create when creating Containers, see
the Parallels Command Line Reference Guide.
Creating a Virtual Machine
The process of creating a new virtual machine includes the following steps:
29
Managing Virtual Machines and Containers
1
Creating a virtual machine configuration. To create a virtual machine configuration, you can use
either the pctl create command or Parallels Management Console.
2
Installing an operating system in the virtual machine. This operation can be performed using
Parallels Management Console only.
3
Installing Parallels Tools in the virtual machine, a set of special utilities that facilitate your work
with virtual machines. This operation can be performed using Parallels Management Console
only.
Note: For detailed information on completing steps 2 and 3, consult the Parallels Management Console
User's Guide.
The example below shows you how to create a new virtual machine configuration using pctl
create:
# pctl create MyVM --distribution win-2008 --location /vz/VMs
Creating the virtual machine...
Generate the VM configuration for win-2008.
The VM has been successfully created.
This will create a virtual machine with the name of MyVM, adjust its configuration for installing the
Windows Server 2008 operating system in it, and place all virtual-machine-related files in the
/vz/VMs directory. Now you can use Parallels Management Console to install Windows Server
2008 OS and Parallels Tools in this virtual machine.
Note: For more information on options you can pass to pctl create when creating virtual machines,
see the Parallels Command Line Reference Guide.
30
Managing Virtual Machines and Containers
Supported Guest Operating Systems
Listed below are the operating systems that you can run in your virtual machines and Containers:
Operating System
Virtual Machine
Container
Windows
Windows 7 with Service Pack 1 (x86, x64)
+
-
Windows Server 2008 R2 with Service Pack 1
(x64)
+
-
Windows Server 2003 R2 with Service Pack 2
(x86, x64)
+
-
Windows Vista with Service Pack 2 (x86, x64)
+
-
Windows XP Professional with Service Pack 2
(x64) and with Service Pack 3 (x86)
+
-
Windows 2000 with Service Pack 4
(x86)
+
-
Red Hat Enterprise Linux 6.1 (x86, x64)
+
+
Red Hat Enterprise Linux 5.6 (x86, x64)
+
+
Red Hat Enterprise Linux 4.8 (x86, x64)
+
-
Fedora 15 (x86, x64)
+
+
Fedora 14 (x86, x64)
-
+
Fedora 13 (x86, x64)
-
+
CentOS 6 (x86, x64)
+
+
CentOS 5.6 (x86, x64)
+
+
CentOS 4.8 (x86, x64)
+
-
SUSE Linux Enterprise Server 11 with Service
Pack 1 (x86, x64)
+
+
SUSE Linux Enterprise Server 10 (x86, x64)
-
+
openSUSE 11.3 (x86, x64)
+
+
openSUSE 11.2 (x86, x64)
-
+
Debian GNU/Linux 6.0 (x86, x64)
+
+
Debian GNU/Linux 5.0 (x86, x64)
+
+
Ubuntu Linux 11.04 (x86, x64)
+
+
Ubuntu Linux 10.10 (x86, x64)
+
+
Ubuntu Linux 10.04 (x86, x64)
+
+
Ubuntu Linux 8.04 (x86, x64)
-
+
+
-
Linux
BSD
FreeBSD 8.1 (x86, x64)
31
Managing Virtual Machines and Containers
FreeBSD 7.3 (x86, x64)
+
-
Oracle Enterprise Linux 5.5 (x86, x64)
+
-
Oracle Enterprise Linux 5.6 (x86, x64)
-
+
Oracle
Compatibility With Parallels Products
You can use virtual machines created in Parallels Server Bare Metal 5.0 with the following Parallels
products:
•
Parallels Desktop 6 for Mac
•
Parallels Server for Mac 4.0
•
Parallels Server for Mac 4.0 Bare Metal Edition
•
Parallels Desktop 4 for Windows and Linux
•
Parallels Workstation 4.0 Extreme
All products must have the latest updates installed.
32
Managing Virtual Machines and Containers
Choosing a Container ID
Every Container has a numeric ID, also known as Container ID, associated with it. The ID is a 32-bit
integer number beginning with zero and unique for a given Parallels server. When choosing an ID
for your Container, please follow the simple guidelines below:
•
ID 0 is used for the Parallels server itself. You cannot and should not try to create a Container
with ID 0.
•
Parallels Server Bare Metal reserves the IDs ranging from 0 to 100. Please do not create
Containers with IDs below 101.
The only strict requirement for a Container ID is to be unique for a particular Parallels server.
However, if you are going to have several computers running Parallels Server Bare Metal, we
recommend assigning different Container ID ranges to them. For example, on server 1 you create
Containers within the range of IDs from 101 to 1000; on server 2 you use the range from 1001 to
2000, and so on. This approach makes it easier to remember on which server a Container has
been created, and eliminates the possibility of Container ID conflicts when a Container migrates
from one Parallels server to another.
Another approach to assigning Container IDs is to follow some pattern of Container IP addresses.
Thus, for example, if you have a subnet with the 10.0.x.x address range, you may want to assign
the 17015 ID to the Container with the 10.0.17.15 IP address, the 39108 ID to the Container with
the 10.0.39.108 IP address, and so on. This makes it much easier to run a number of Parallels
utilities eliminating the necessity to check up the Container IP address by its ID and similar tasks.
You can also think of your own patterns for assigning Container IDs depending on the configuration
of your network and your specific needs.
Before you decide on a new Container ID, you may want to make sure that no Container with this
ID has yet been created on the server. The easiest way to check this is to run the following
command:
# vzlist -a 101
Container not found
This output shows that Container 101 does not exist on the server; otherwise it would be present in
the list.
WARNING! When deciding on a Container ID, do not use IDs that were once assigned to
Containers unless you are sure that no data belonging to the old Containers remains on the server.
Otherwise, the administrator of the newly-created Container may get access to this data—that is,
to the backups of the old Container, its logs, statistics, and so on.
33
Managing Virtual Machines and Containers
Choosing OS EZ Template
Before starting to create a Container, you shall decide on which OS EZ template your Container will
be based. There might be several OS EZ templates installed on the server and prepared for the
Container creation; use the vzpkg list command to find out what OS EZ templates are available
on your system:
# vzpkg list -O
redhat-el5-x86
fedora-core-13-x86
2010-05-21 23:59:44
2010-12-11 12:45:52
The -O option passed to the vzpkg list command allows you to list only OS EZ templates
installed on the server. As you can see, the redhat-el5-x86 and fedora-core-13-x86 OS
EZ templates are currently available on the server. The time displayed next to OS EZ templates
indicates when these templates were cached.
You can also use the --with-summary option to display brief information on the installed OS EZ
templates:
# vzpkg list -O --with-summary
redhat-el5-x86
:Red Hat Enterprise Linux v.5 Server EZ OS template
fedora-core-13-x86 :Fedora 13 EZ OS template
For detailed information on the vzpkg list command, consult the Parallels Command Line
Reference Guide.
Performing Initial Configuration
Before starting your newly created virtual machine and Container, you first need to configure it. This
section describes the main configuration steps for virtual machines and Containers.
34
Managing Virtual Machines and Containers
Configuring Network Settings
To make virtual machines and Containers accessible from the network, you need to assign valid IP
addresses to them and configure DNS servers. The session below illustrates setting these
parameters for the MyVM virtual machine and Container 101:
•
#
#
#
#
Assigning IPv4 and IPv6 addresses:
pctl
pctl
pctl
pctl
set
set
set
set
MyVM --device-set net0 --ipadd 10.0.186.100/24
MyVM --device-set net0 --ipadd 1fe80::20c:29ff:fe01:fb07
101 --ipadd 10.0.186.101/24 --save
101 --ipadd fe80::20c:29ff:fe01:fb08 --save
net0 in the commands above denotes the network card in the VM virtual machine to assign the
IP address to. You can view all network cards of a virtual machine using the pctl list
VM_name -i command.
•
Setting DNS server addresses:
# pctl set MyVM --nameserver 192.168.1.165
# pctl set 101 --nameserver 192.168.1.165 --save
When running commands for a Container, you need to use the –-save flag to save the specified
parameters to the Container configuration file. If you omit this option, the applied values will be valid
only until the Container shutdown. Omit the --save flag when executing commands for a virtual
machine; the parameters are automatically saved to the virtual machine configuration file without
this flag.
Notes:
1. You can configure the network settings only for virtual machines that have Parallels Tools installed.
2. To assign network masks to Containers operating in the venet0 networking mode, you must set the
USE_VENET_MASK parameter in the /etc/vz/vz.conf configuration file to yes.
35
Managing Virtual Machines and Containers
Setting Passwords for Virtual Machines and Containers
In Parallels Server Bare Metal, you can use the --userpasswd option of the pctl set
command to create new accounts in your virtual machines and Containers directly from the
Parallels server. The created account can then be used to log in to the virtual machine and
Container. The easiest way of doing it is to run this command:
# pctl set MyVM --userpasswd user1:2wsx123qwe
This command creates the user1 account in the MyVM virtual machine and sets the 2wsx123qwe
password for it. Now you can log in to the MyVM virtual machine as user1 and administer it in the
same way you would administer a standalone server: install additional software, add users, set up
services, and so on.
The pctl set command can also be used to change passwords for existing accounts in your virtual
machines and Containers. For example, to change the password for user1 in the MyVM virtual
machine to 0pi65jh9, run this command:
# pctl set MyVM --userpasswd user1:0pi65jh9
When setting passwords for virtual machines and Containers, keep in mind the following:
•
You can use manage user accounts only inside virtual machines that have Parallels Tools
installed.
•
You should use passwords that meet the minimum length and complexity requirements of the
respective operating system. For example, for Windows Server 2008, a password must be
more than six characters in length and contain characters from three of the following categories:
uppercase characters, lowercase characters, digits, and non-alphabetic characters.
•
You should not create accounts with empty passwords for virtual machines and Containers
running Linux operating systems.
Setting Startup Parameters
The pctl set command allows you to define the onboot startup parameter for virtual machines
and Containers. Setting this parameter to yes makes your virtual machine and Container
automatically boot at the physical server startup. For example, to enable Container 101 and the
MyVM virtual machine to automatically start on your server boot, you can execute the following
commands:
•
For Container 101:
# pctl set 101 --onboot yes --save
Saved parameters for Container 101
•
For the MyVM virtual machine:
# pctl set MyVM --onboot yes
Notice that the onboot parameter will have effect only on the next server startup.
36
Managing Virtual Machines and Containers
Starting, Stopping, and Querying Status of a
Virtual Machine and Container
After a Parallels virtual machine and Container has been created, it can be managed like an
ordinary computer.
Starting a Virtual Machine and Container
You can use the pctl start command to start your virtual machines and Containers:
•
To start Container 101:
# pctl start 101
Starting the Container ...
•
To start a virtual machine with the name of MyVM:
# pctl start MyVM
Starting the VM ...
Stopping a Virtual Machine and Container
The pctl stop command is used to stop your virtual machines and Containers:
•
To stop Container 101:
# pctl stop 101
Stopping the Container ...
•
To stop a virtual machine with the name of MyVM:
# pctl stop MyVM
Stopping the VM ...
Checking the Status of a Virtual Machine and Container
Depending on whether you want to check the status of a Container or a virtual machine, you can
use the following commands:
•
pctl status to check the Container status:
# pctl status 101
VEID 101 exist mounted running
•
pctl list to check the virtual machine status:
# pctl list MyVM
stopped 10.12.12.121 MyVM
You can also get more detailed information on a virtual machine by specifying the -i option
after pctl list.
37
Managing Virtual Machines and Containers
Restarting a Virtual Machine and Container
Sometimes, you may need to restart a virtual machine and Container. To do this, use the following
commands:
•
To restart a Container, use the pctl restart command:
# pctl restart 101
Stopping Container ...
Container was stopped
Container is unmounted
Starting Container ...
Container is mounted
Adding IP address(es): 10.0.186.101
Container start in progress...
•
To restart a virtual machine, use the pctl reset command:
# pctl reset MyVM
Listing Virtual Machines and Containers
To get an overview of the virtual machines and Containers existing on the physical server and to get
additional information about them - their IP addresses, hostnames, current resource consumption,
and so on - use the pctl list command. In the most general case, you can get a list of all virtual
machines and Containers by issuing the following command:
# pctl list -a
STATUS
IP_ADDR
started 10.10.1.101
stopped 10.10.100.1
NAME
101
MyVM
The -a option tells the pctl list command to output both running and stopped virtual
machines and Containers. By default, only running virtual machines and Containers are shown. The
default columns inform you of the Container IDs and virtual machine names, the virtual machine
and Container status and IP addresses. This output can be customized as desired by using pctl
list command line options. For example:
# pctl list -a -o name,ctid -a
NAME
ID
101
MyVm
{b8cb6d99-1af1-453d-a302-2fddd8f86769}
This command displays only the names and IDs of the virtual machines and Containers existing on
the physical server. The full list of the pctl list command options for virtual machines and
Containers is available in the Parallels Command Line Reference Guide.
38
Managing Virtual Machines and Containers
Storing Extended Information on a Virtual
Machine and Container
Sometimes, it may be difficult to remember the information on certain virtual machines and
Containers. The probability of this increases together with the number of virtual machines and
Containers and with the time elapsed since their creation. Parallels Server Bare Metal allows you to
set the description of any virtual machine and Container on the physical server and view it later on,
if required. The description can be any text containing any virtual machine and Container-related
information. For example, you can include the following in the virtual machine and Container
description:
•
the owner of the virtual machine and Container
•
the purpose of the virtual machine and Container
•
the summary description of the virtual machine and Container
Let us assume that you are asked to create a virtual machine for a Mr. Johnson who is going to use
it for hosting the MySQL server. So, you create the MyVM virtual machine and, after that, execute
the following command on the physical server:
# pctl set MyVM --description "MyVM
> owner - Mr. Johnson
> purpose - hosting the MySQL server" The VM has been successfully configured.
This command saves the following information related to the virtual machine: its name, owner, and
the purpose of its creation. At any time, you can display this information by issuing the following
command:
# pctl list -o description MyVM
MyVM
owner - Mr. Johnson
purpose - hosting the MySQL server
When working with virtual machine and Container descriptions, keep in mind the following:
•
You can use any symbols you like in the virtual machine and Container description (new lines,
dashes, underscores, spaces, etc.).
•
If the virtual machine and Container description contains one or more spaces or line breaks (as
in the example above), it must be put in single or double quotes.
•
As distinct from a virtual machine and Container name and ID, a description cannot be used for
performing virtual machine and Container-related operations (e.g. for starting or stopping a
virtual machine and Container) and is meant for reference purposes only.
39
Managing Virtual Machines and Containers
Copying a Virtual Machine and Container Within
the Server
Parallels Server Bare Metal allows you to create a complete copy of a particular virtual machine and
Container (in respect of all the virtual machine and Container data and resources parameters), or a
clone. This saves your time because you do not have to think of setting up the virtual machine and
Container configuration parameters and the like. Moreover, you can create a number of virtual
machine and Container clones at a sitting.
In Parallels Server Bare Metal-based systems, you can use the following commands to copy a
virtual machine and Container within the given physical server:
•
vzmlocal to clone a Container. For example, you can create Container 111 and make it be a
complete copy of Container 101 by running this command:
# vzmlocal -C 101:111
Moving/copying Container#101 -> Container#111, [], [] ...
...
Successfully completed
You can clone both running and stopped Containers.
•
pctl clone to clone a virtual machine. For example, you can create a clone of the MyVM
virtual machine and assign the Cloned_VM name to it as follows:
# pctl clone MyVM --name ClonedVM
Clone the MyVM VM to the VM ClonedVM...
The VM has been successfully cloned.
You can create clones of stopped virtual machines only.
Checking the Cloned Virtual Machine and Container
To check that your virtual machine and Container has been successfully moved, run this command:
# pctl list -a
STATUS
IP_ADDR
stopped
10.0.10.101
stopped
10.0.10.101
stopped
10.0.10.115
stopped
10.0.10.115
NAME
101
111
MyVM
ClonedVM
As you can see from the example above, the clones of Container 101 (Container 111) and the
MyVM virtual machine (ClonedVM) have been successfully created. However, before starting to use
the clones, you should assign different IP addresses to them which are currently identical to those
of Container 101 and MyVM. Refer to Performing Initial Configuration (p. 34) to learn how you can
do it.
Note: If you are cloning a running Container, the created clone is stopped to prevent an IP address
conflict.
40
Managing Virtual Machines and Containers
Configuring the Default Directories
When cloning a virtual machine and Container, you can also override the following default
directories:
•
/vz/dest_VM_Name.pvm storing the files of a cloned virtual machine (where
dest_VM_Name denotes the name of the resulting virtual machine). For example, for the
ClonedVM virtual machine, this directory is /vz/ClonedVM.pvm. To store the files of the
ClonedVM virtual machine in a custom directory, you can run the following command:
# pctl clone MyVM --name ClonedVM --location /vz/VM_directory
In this case all virtual machine files will be placed to the /vz/VM_directory directory. Notice
that the specified directory must exist on the server; otherwise, the command will fail.
•
/vz/private/<dest_CTID> and /vz/root/<dest_CTID> defining the Container private
area and root paths, respectively (where <dest_CTID> denotes the ID of the resulting
Container). In the case of Container 111, these paths are /vz/private/111 and
/vz/root/111. To define custom private area and root paths for Container 111, you can
execute the following command:
# vzmlocal -C 101:111:/vz/private/dir_111/:/vz/root/ct111
Moving/copying Container#101 -> Container#111, [], [] ...
Syncing private area '/vz/private/101'->'/vz/private/dir_111'
...
Successfully completed
# ls /vz/private
1 101 dir_111
# ls /vz/root
1 101 ct111
41
Managing Virtual Machines and Containers
Suspending a Virtual Machine and Container
Parallels Server Bare Metal allows you to suspend a running virtual machine and Container on the
physical server by saving its current state to a special file. Later on, you can resume the virtual
machine and Container and get it in the same state the virtual machine and Container was at the
time of its suspending. Suspending your virtual machines and Containers may prove useful, for
example, if you need to restart the physical server, but do not want to:
•
quit the applications currently running in the virtual machine and Container
•
spend much time on shutting down the guest operating system and then starting it again
You can use the pctl suspend command to save the current state of a virtual machine and
Container. For example, you can issue the following command to suspend the MyVM virtual
machine:
# pctl suspend MyVM
Suspending the VM...
The VM has been successfully suspended.
At any time, you can resume the MyVM virtual machine by executing the following command:
# pctl resume MyVM
Resuming the VM...
The VM has been successfully resumed
Once the restoration process is complete, any applications that were running in the MyVM virtual
machine at the time of its suspending will be running again and the information content will be the
same as it was when the virtual machine was suspended.
42
Managing Virtual Machines and Containers
Running Commands in a Virtual Machine and
Container
Parallels Server Bare Metal allows you to execute arbitrary commands inside virtual machines and
Containers by running them on the physical server, i.e. without the need to log in to the respective
virtual machine and Container. For example, this can be useful in these cases:
•
If you do not know the virtual machine and Container login information, but need to run some
diagnosis commands to verify that it is operational.
•
If network access is absent for a virtual machine and Container.
In both these cases, you can use the pctl exec command to run a command inside the
respective virtual machine and Container. The session below illustrates the situation when you run
the stopped SSH daemon inside a Linux virtual machine with the name of My_Linux:
# pctl exec My_Linux /etc/init.d/sshd status
sshd is stopped
# pctl exec My_Linux /etc/init.d/sshd start
Starting sshd:[OK]
# pctl exec My_Linux /etc/init.d/sshd status
sshd (pid 26187) is running...
Notes:
1. You can use the pctl exec command only inside virtual machines that have Parallels Tools installed.
2. The pctl exec command is executed inside a virtual machine and Container from the / directory
rather than from the /root one.
43
Managing Virtual Machines and Containers
Deleting a Virtual Machine and Container
You can delete a virtual machine and Container that is not needed anymore using the pctl
delete command. Notice that you cannot delete a running or mounted virtual machine and
Container. The example below illustrates deleting Container 101 and the MyVM virtual machine:
Deleting Container 101
# pctl delete 101
Deleting Container private area: /vz/private/101
Container is currently mounted (unmount first)
# pctl stop 101
Stopping Container...
Container was stopped
Container is unmounted
# pctl delete 101
Deleting Container private area: /vz/private/101
Container private area was deleted
Deleting the MyVM virtual machine:
# pctl delete MyVM
Deleting the VM...
VM is currently running
# pctl stop MyVM
Stopping the VM...
VM was stopped
# pctl delete MyVM
Deleting the VM...
Container was deleted
Managing Virtual Machine and Container
Backups
A regular backing up of the existing virtual machines and Containers is essential for any physical
server reliability. In Parallels Server Bare Metal, you can use the following utilities to back up and
restore your virtual machines and Containers:
•
pctl
•
pbackup
•
prestore
Detailed information on these utilities is provided in the following subsections.
44
Managing Virtual Machines and Containers
Backups Overview
Parallels Server Bare Metal backup utilities deal with three kinds of servers:
•
Source Server. This is the server where virtual machines and Containers are hosted during their
backing up.
•
Backup Server. This is the server where virtual machine and Container backups are stored. A
Backup Server can be any server running the Parallels Server Bare Metal software and having
sufficient space for storing virtual machine and Container backups.
•
Destination Server. This is the server where virtual machine and Container backups are
restored.
45
Managing Virtual Machines and Containers
These servers are singled out by their functionality only. In reality, one and the same physical server
can perform two or even three functions. Usually, the Source and Destination Servers are
represented by one and the same server because you will likely want the virtual machines and
Containers you back up to be restored to their original server. However, setting up a dedicated
Backup Server is recommended.
Creating Consistent Backups of Virtual Machines
Parallels Server Bare Metal allows you to back up both running and stopped virtual machines.
However, to create a consistent backup of a running virtual machine, the virtual machine must meet
the following requirements:
•
Have Parallels Tools installed.
•
Run one of the following operating systems:
Windows operating systems
• Windows Server 2003
• Windows Server 2008
• Windows Vista
• Windows 7
Linux operating systems
• Suse, version 9.0 and higher
• RHEL, version 4.0 and higher
• CentOS, version 4.0 and higher
• Fedora Core, version 3 and higher
• Debian, version 3.1 and higher
• Ubuntu, version 4.10 and higher
46
Managing Virtual Machines and Containers
Using pctl backup and pctl restore
This section describes how to perform the basic backup-related operations using the pctl utility.
Creating a Virtual Machine and Container Backup
You can use the pctl backup command to back up virtual machines and Containers. This
command is executed on the Source Server and can store the created virtual machine and
Container backup on both the Source and Backup Servers. When creating a backup on the Source
Server, you only need to specify the name of the virtual machine and Container to back up. For
example, you can execute the following command to back up the MyVM virtual machine and store
its backup archive on the Source Server:
# pctl backup MyVM
Backing up the VM MyVM
Operation progress 100%
The virtual machine has been successfully backed up with backup ID {746dba2a-3b10-4ced9dd6-76a2blcl4a69}
The command output informs you that the virtual machine backup has been successfully created
and assigned ID 746dba2a-3b10-4ced-9dd6-76a2blcl4a69. You can use this ID when
managing the backup archive (e.g. remove the backup).
At the same time, you can run the following command to back up the MyVM virtual machine and
store its backup archive on the Backup Server with the IP address of 129.129.10.10:
# pctl backup MyVM -s root:1qaz2wsx@129.129.10.10
root:1qaz2wsx before the Destination Server IP address denotes the root credentials used to
log in to this server. If you do not specify these credentials, you will be asked to do so during the
command execution.
By default, all newly created backups are placed to the following directories:
•
/vz/backups for Containers
•
/vz/vmprivate/backups for virtual machines
For Containers, you can configure the default directory by changing the value of the BACKUP_DIR
parameter in the /etc/vzbackup.conf configuration file. In turn, to set the default backup
directory for virtual machines, you should use the prlsrvctl set command.
Notes:
1. A Backup Server can be any server running the Parallels Server Bare Metal software and having
sufficient space for storing virtual machine and Container backups.
2. For more information on the options you can pass to pctl backup, refer to the Parallels Server Bare
Metal 5.0 Command Line Reference Guide.
47
Managing Virtual Machines and Containers
Listing the Existing Backups
You can use the pctl backup-list command to view the backups existing on the physical
server. For example:
# pctl backup-list
ID Backup_ID
Node
Date
Type Size
{c1dee22f...} {209d54a0...} test.com 2011-05-30 10:19:32 f
411566405
[The ID and Backup ID are reduced for better readability.]
This command lists the backups existing on the Source Server. If you want to list the backups on
the Backup Server, you need to specify the IP address of this server.
The command output shows that currently only one backup exists on the Source Server. This
backup was assigned the ID of c1dee22f-8667-4870-9e11-278f1398eab0 (the full ID is
skipped in the command output). The information on the backup is presented in the following table:
Column Name
Description
ID
The ID uniquely identifying the virtual machine and Container.
Backup ID
The ID assigned to the backup archive. You need to specify this ID when
performing any backup-related operations.
Node
The hostname of the physical server storing the backup archive.
Date
The date and time when the backup archive was created.
Type
The backup type. Currently, you can create two types of backups:
Size
•
A full backup indicated by f.
•
An incremental backup indicated by i and containing only the
files changed since the previous full or incremental backup. This
is the default backup type.
The size of the backup archive, in bytes.
Removing a Virtual Machine and Container Backup
At any time, you can remove a backup that you do not need any more using the pctl backupdelete command. To do this, you need to specify the ID of the backup to remove and the ID of
the respective virtual machine and Container. If you do not know these IDs, use the pctl
backup-list and check the ID and Backup ID columns. For example:
# pctl backup-list
ID Backup_ID
Node
Date
Type Size
{c1dee22f...} {209d54a0...} test.com 2011-05-30 10:19:32 f
411566405
[The ID and Backup ID are reduced for better readability.]
# pctl backup-delete c1dee22f-8667-4870-9e11-278f1398eab0 -t 209d54a0-e3b8-4a03-9ca8d4cc7a2a27ca
Delete the VM backup
The VM backup has been successfully removed.
You can also specify the virtual machine and Container name instead of its ID:
# pctl backup-delete MyVM -t 209d54a0-e3b8-4a03-9ca8-d4cc7a2a27ca
48
Managing Virtual Machines and Containers
If you have several backups of a particular virtual machine and Container and want to delete them
all at once, indicate only the virtual machine and Container name or ID:
# pctl backup-delete MyVM
This command removes all backups of the MyVM virtual machine from the local Backup Server. To
remove backups stored remotely, you also need to specify the IP address of the remote Server:
# pctl backup-delete MyVM -s root:1qaz2wsx@129.129.10.10
Restore a Virtual Machine and Container
To restore a backup of a virtual machine and Container, you can use the pctl restore
command. This command supports restoring backups to the Source Server only. For example, to
restore a backup of the MyVM virtual machine stored on the Backup Server with the IP address of
10.10.100.1, you can run this command on the Source Node:
# pctl restore MyVM -s root:1qaz2wsx@10.10.100.1
If you have two or more backups of the MyVM virtual machine, the latest backup is restored. If you
want to restore a particular virtual machine and Container backup, you need to specify the ID of this
backup. You can use the pctl backup-list command to list the existing backups and the IDs
assigned to them:
# pctl backup-list -s root:1qaz2wsx@10.10.100.1
ID Backup_ID
Node
Date
Type Size
{c1dee22f...} {209d54a0...} test.com 2011-05-30 10:19:32 i
11566405
{c1dee22f...} {24a3011c...} test.com 2011-05-21 11:12:35 f
356798701
[The ID and Backup ID are reduced for better readability.]
You can now indicate the desired ID after the -t option to tell pctl backup to restore this
particular backup. For example, to restore the backup for the virtual machine with the ID of
c1dee22f-8667-4870-9e11-278f1398eab0 that was created on the 21st of May, you can
execute this command:
# pctl restore -t {24a3011c-8667-4870-9e11-278f1398eab0} -s root:1qaz2wsx@10.10.100.1
Note: Virtual machines created on servers running Parallels Server Bare Metal 5.0 cannot be restored on
servers with Parallels Server 4 Bare Metal.
49
Managing Virtual Machines and Containers
Using pbackup and prestore
Along with pctl, you can use the following utilities to create and manage backups of your virtual
machines and Containers:
•
pbackup. This utility is used to create backups of individual virtual machines and Containers or
entire Parallels servers.
•
prestore. This utility is used to manage the existing backups of virtual machines and
Containers.
Backing Up Virtual Machines and Containers
The pbackup utility is run on the Backup Server connecting via SSH to the Parallels server and
backing up one or more virtual machines and Containers on this server. By default, the created
Container backups are placed to the /vz/backups directory and the backups of virtual machines
are put to the /vz/vmprivate/backups directory. You can change the default backup
directories
•
for Containers, by editing the /etc/vzbackup.conf file
•
for virtual machines, by using the --backup-path option of the prlsrvctl set command
Let us assume that you want to back up the entire Parallels server (that is, all virtual machines and
Containers on this server) with the test.com hostname. In this case, you can run the following
command on the Backup Server:
# pbackup test.com
During the command execution, you will be asked to provide the test.com credentials. After
doing so, the command will back up all virtual machines and Containers on the test.com and put
•
all backed up Containers to the Backup Server
•
all backed up virtual machines to the Source Server
To save the backed up virtual machines also on the Backup Server, you should additionally specify
the -n option. This option is used to indicate the IP address or hostname of the Backup Server and
its credentials:
# pbackup -n root:7ujn6yhb@192.168.10.199 test.com
If you wish to back up not all, but specific virtual machines and Containers from the specified
server, use the –e or –x switches (to include or exclude the specified virtual machines and
Containers, respectively). For example:
# pbackup -n root:7ujn6yhb@192.168.10.199 test.com -e 101 MyVM
In this session, only Container 101 and the MyVM virtual machine residing on the Source Server with
the test.com hostname will be included in the backup, and their backups will be stored on the
Backup Server.
50
Managing Virtual Machines and Containers
Notes:
1. A Backup Server can be any server running the Parallels Server Bare Metal software and having
sufficient space for storing virtual machine and Container backups.
2. For the full list of configuration parameters and command line options for pbackup, consult the
Parallels Server Bare Metal 5.0 Command Line Reference Guide.
Restoring Backups
To restore any individual virtual machines and Containers or entire Parallels servers, you may want
to view first the information about them. This can be done using the prestore -l command:
# prestore -l -n test.com test.com
root@test.com's password:
...
Backups for node test.com:
ID Backup_ID
Node
Date
Type Size
101 2011-05-...
test.com 2011-05-30 09:42:19 f
18721280
{cd91b90b...} {4ef87485...} test.com 2011-05-16 17:15:47 f
92617398
[The ID and Backup ID are reduced for better readability.]
The command output shows that currently only two backups exist for the test.com server on the
Backup Server. If you omit the -n test.com option, the command will list:
•
all Container backups for the test.com server stored on the Backup Server
•
all virtual machine backups for the test.com server stored on the test.com server
The information on the backups is presented in the following table:
Column Name
Description
ID
The ID uniquely identifying the virtual machine and Container.
Backup ID
The ID assigned to the backup archive. You need to specify this ID when
performing any backup-related operations.
Node
The hostname of the Source Server.
Date
The date and time when the backup archive was created.
Type
The backup type. Currently, you can create two types of backups:
Size
•
A full backup indicated by f.
•
An incremental backup indicated by i and containing only the
files changed since the previous full or incremental backup. This
is the default backup type.
The size of the backup archive, in bytes.
To restore Container 101 and the {cd91b90b-469d-42c6-acf4-fefee09cfa61} virtual
machine, run this command:
# prestore -n test.com -e 101 {cd91b90b-469d-42c6-acf4-fefee09cfa61}
This command will restore the Container and the virtual machine to their Source Server.
51
Managing Virtual Machines and Containers
You can also use the -d option to restore Container 101 to a Parallels server other than the Source
Node. For example, this command
# prestore -d 192.168.10.199 test.com -e 101
restores Container 101 to the Destination Server with IP address 192.168.10.199. If you want to
restore all Containers backups for the test.com Parallels server, just skip the -e option.
Notes:
1. The current version of Parallels Server Bare Metal supports restoring virtual machines to the Source
Server only.
2. Virtual machines created on servers running Parallels Server Bare Metal 5.0 cannot be restored on
servers with Parallels Server 4 Bare Metal.
3. The prestore utility can also manage (list, restore, etc.) backups created using the pctl backup
command. However, you are highly recommended to use the same utility (either pctl or prestore)
during the life cycle of a particular backup.
4. For the full list of command line options for prestore, refer to the Parallels Server Bare Metal 5.0
Command Line Reference Guide.
52
Managing Virtual Machines and Containers
Configuring Passwordless Access to the Source Node
You need to provide the Source Node credentials each time you execute the pbackup and
prestore commands. However, you can allow these utilities to log in to the Source Node without
having to enter the root password. To do this, you need to provide each Source Node with
authorized public SSH RSA keys:
1
Log in to the Backup Server as root, and generate a pair of SSH keys - public and private:
# ssh-keygen -t rsa
Generating public/private rsa key pair.
Enter file in which to save the key (/root/.ssh/id_rsa):
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /root/.ssh/id_rsa.
Your public key has been saved in /root/.ssh/id_rsa.pub.
The key fingerprint is:
c6:19:a8:2c:67:31:15:e6:30:23:2b:8a:b0:63:77:8f root@dhcp-130.parallels.com
Note that you must leave an empty passphrase in the above procedure. The private key is
saved by default in /root/.ssh/id_rsa, and the public key is saved in
/root/.ssh/id_rsa.pub.
2
Transfer your public key to the /root/.ssh directory on each Source Node (use some
intermediary name for the file not to overwrite the corresponding file on the Source Node):
# scp /root/.ssh/id_rsa.pub root@dhcp-129.parallels.com:/root/.ssh/temp_name
The authenticity of host 'dhcp-129.parallels.com (192.168.1.129)' can't be established.
RSA key fingerprint is 01:fc:b6:e9:26:40:1f:1a:41:5f:7a:fb:cf:14:51.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'dhcp-129.parallels.com,192.168.1.129' (RSA) to the list of
known hosts.
root@dhcp-129.parallels.com's password:
id_rsa.pub
100% |*****************************|
235
00:00
3
Add the contents of the transferred file to the authorized_keys file in this very directory on
the Source Node. To do this, log in to the Source Node, change to the /root/.ssh directory,
and issue the following command:
# cat temp_name >> authorized_keys
Now the pbackup/prestore utilities should be able to log in to the Source Node as root
without having to provide the root password.
53
Managing Virtual Machines and Containers
Migrating Virtual Machines and Containers
The Parallels physical server is the system with higher availability requirements in comparison with a
typical system. If you are running your company mail server, file server, and web server in different
virtual machines and Containers on one and the same physical server, then shutting it down for
hardware upgrade will make all these services unavailable at once. To facilitate hardware upgrades
and load balancing between several Parallels servers, the Parallels Server Bare Metal software
provides you with the ability to migrate virtual machines and Containers from one physical server to
another.
Parallels Server Bare Metal is shipped with a special utility—pmigrate—allowing you to perform
different types of migration. Using this utility, you can migrate
•
Containers from one physical server to another
•
Parallels virtual machines from one physical server to another
•
a Container to a Parallels virtual machine
•
a Parallels virtual machine to a Container
•
a physical server to a virtual machine and Container
•
Xen virtual machines to Parallels virtual machines
All these operations are described in the following subsections.
54
Managing Virtual Machines and Containers
General Migration Requirements
Before deciding on the type of migration to perform, make sure that the source computer (i.e. the
computer that you will migrate or that stores the virtual machine and Container before its migration)
and the destination computer (i.e. the computer that runs Parallels Server Bare Metal and that will
host the resulting virtual machine and Container) meet the requirements below.
Requirements for the source computer
The source computer can be a physical computer, a virtual machine, or a Container. The software
requirements for source computers are given in the following table:
Operating System
Physical Computer
Virtual Machine
Container
Windows 7 (x32, x64)
+
+
-
Windows Server 2003 (x32, x64)
+
+
+
Windows Server 2008 (x32, x64)
+
+
+
Windows 2000 Server (x32)
+
+
-
Windows XP (x32, x64)
+
+
-
Windows Vista (x32, x64)
+
+
-
Red Hat Enterprise Linux 5 (x32,
x64)
+
+
+
Red Hat Enterprise Linux 4 (x32,
x64)
+
+
+
CentOS 5 (x32, x64)
+
+
+
CentOS 4 (x32, x64)
+
+
+
Fedora 11 (x32, x64)
+
+
+
Fedora 10 (x32, x64)
+
+
+
SUSE Linux Enterprise Server 10
(x32, x64)
+
+
+
Debian GNU/Linux 5 (x32, x64)
+
+
+
Debian GNU/Linux 4 (x32, x64)
+
+
+
Ubuntu Linux 9.04 Server (x32, x64) +
+
+
Ubuntu Linux 8.10 Server (x32, x64) +
+
+
Windows
Linux
Note: In the current version of Parallels Server Bare Metal, you cannot migrate Containers running
Windows Server 2008 to virtual machines.
55
Managing Virtual Machines and Containers
Requirements for the destination server
The destination server must meet the following requirements:
•
Has enough hard disk space to store the resulting virtual machine and Container.
•
Has enough memory and CPU power to run the resulting virtual machine and Container.
•
Has a stable network connection with the source server.
Migrating Virtual Machines and Containers Between Parallels Servers
In Parallels Server Bare Metal, you can choose one of the following ways to migrate virtual
machines and Containers:
•
Migrating virtual machines and Containers using the standard migration technology.
•
Migrating virtual machines and Containers using the zero-downtime migration technology.
Both ways of migrating virtual machines and Containers are described in the following subsections
in detail.
56
Managing Virtual Machines and Containers
Standard Migration
Using the standard migration technology, you can move
•
stopped, running, and suspended Containers
•
stopped and suspended virtual machines
Standard migration includes copying all files of a virtual machine or Container from one Parallels
server to another and does not differ from copying a number of files from one server to another
over the network. For a running Container, the migration procedure is a bit more complicated and
is described below:
1
Once you start the migration, all Container data are copied to the destination server. During this
time, the Container on the source server continues running.
2
The Container on the source server is stopped.
3
The Container data copied to the destination server are compared with those on the source
server, and if any files were changed during the first migration step, they are copied to the
destination server again and rewrite the outdated versions.
4
The Container on the destination server is started.
There is a short downtime needed to stop the Container on the source server, copy the changed
data to the destination server, and start the Container on the destination server.
Note: Before the migration, you may need to detach the Container from its caches. For more information
on cached files, see the Cleaning Up Containers subsection (p. 122).
Migrating Containers
The following session moves Container 101 from the source server to the destination server
ts7.test.com:
# pmigrate c 101 c root:1qasdeqw3@ts7.test.com/101
root@ts7.test.com's password:
vzmsrc: Connection to destination server (ts7.test.com) is successfully established
...
Successfully completed
The c option in the command above tells pmigrate that you are moving a Container to a
Container. If you do not indicate the credentials to log in to the destination server, you will need to
do so during the migration.
Important! For the command to be successful, a direct SSH connection (on port 22) must be allowed
between the source and destination servers.
57
Managing Virtual Machines and Containers
By default, after the migration process is completed, the Container private area and configuration
file are renamed on the source server by receiving the .migrated suffix. However, if you want the
Container private area on the source server to be removed after the successful Container migration,
you can override the default pmigrate behavior by changing the value of the REMOVEMIGRATED
variable in the Parallels Server Bare Metal global configuration file (/etc/vz/vz.conf) to yes or
by using the –r yes switch with the pmigrate command.
Migrating Virtual Machines
In turn, to migrate a virtual machine from the source server to ts7.test.com, you need just to
specify v instead of c and the name of the resulting virtual machine instead of Container ID 101:
# pmigrate v MyVM v ts7.test.com/MyVM
Migrate the VM MyVM to test.com
root@ts7.test.com's password:
Operation progress 100%
The VM has been successfully migrated.
This command moves the MyVM virtual machine from the local server to the destination server
ts7.test.com. Once the migration is complete, the original virtual machine is removed from the
source server. However, you can use the --keep-src option to leave the original virtual machine
intact.
For virtual machines, pmigrate also supports the migration from a remote Parallels server to the
local one:
# pmigrate v ts7.test.com/MyVM v localhost
root@ts7.test.com's password:
Migrate the VM MyVM to localhost
Operation progress 100%
The VM has been successfully migrated.
This command moves the MyVM virtual machine from the ts7.test.com server to the local
server.
Note: For more information on options that you can pass to pmigrate, refer to the Parallels Server Bare
Metal 5.0 Command Line Reference.
58
Managing Virtual Machines and Containers
Zero-Downtime Migration
Using the zero-downtime migration technology, you can migrate paused and running virtual
machines and running Containers from one Parallels server to another with zero downtime. The
zero-downtime migration technology has the following main advantages over the standard one:
•
The migration time is greatly reduced. In fact, the migration eliminates the service outage or
interruption for end users.
•
The process of migrating a virtual machine or Container to another Parallels server is
transparent for you and the Container applications and network connections. This means that
no modifications of system characteristics and operational procedures inside the Container are
performed on the source and destination servers.
•
The virtual machine or Container is restored on the destination server in the same state as it
was at the beginning of the migration.
•
You can move virtual machines and Containers running applications that you do not want to be
rebooted during the migration.
Note: Zero-downtime migration is not supported for virtual machines and Containers that have open
sessions established with the pctl enter command.
Migration requirements and restrictions
When performing a zero-downtime migration, take into account the requirements and restrictions
below:
•
Before performing zero-downtime migration, it is recommended to synchronize the system time
on the source and destination servers, for example, by means of NTP (http://www.ntp.org). The
reason for this recommendation is that some processes running in virtual machines and
Containers might rely on the system time being monotonic and thus might behave
unpredictably if they see an abrupt step forward or backward in the time once they find
themselves on the new server with different system clock parameters.
•
Your network must support data transfer rates of at least 1 Gb/s.
•
The source and destination servers must belong to the same subnetwork.
•
The CPUs on the source and destination servers must be manufactured by the same vendor,
and the CPU capabilities on the destination server must be the same or exceed those on the
source server.
•
virtual machine and Container disks can be located on local disks, shared NFS and GFS2
storages, and ISCSI raw devices.
Migration process overview
The process of migrating virtual machines and Containers using the zero-downtime migration
technology includes the following main steps:
59
Managing Virtual Machines and Containers
1
Once you start the migration, Parallels Server Bare Metal checks whether the destination server
meets all migration requirements and the virtual machine or Container can be migrated to this
server.
2
All virtual memory and disks of the virtual machine or Container are migrated to the destination
server.
3
The virtual machine or Container on the source server is suspended.
4
The changed memory pages and virtual disk blocks, if any, are migrated to the destination
server.
5
The virtual machine or Container is resumed on the destination server.
The virtual machine or Container continues running during steps 1 and 2 and is not available to the
end user during steps 3-5. But since the amount of memory pages and virtual disk blocks changed
during step 2 is small, the service outage time for the end user is almost imperceptible.
Migrating virtual machines and Containers
Depending on whether you are migrating a virtual machine or Container, the command-line options
you pass to the pmigrate slightly differ. For example, you can migrate Container 101 from the
local server to the destination server destserver.com by executing the following command on
the local server:
# pmigrate c 101 c --online destserver.com
Enter password:
Connection to destination server (192.168.1.57) is successfully established
...
Successfully completed
At the same time, to migrate the MyVM virtual machine to the same destination server
destserver.com, you can run this command on the local server:
# pmigrate v MyVM v destserver.com
Migrate the VM MyVM to test.com
...
The VM has been successfully migrated.
As you can see, to migrate a virtual machine, you skip the --online option and use the v option
to specify that you are migrating a virtual machine.
Notes:
1. For more information on options you can use with the pmigrate utility, see the Parallels Server Bare
Metal 5.0 Command-Line Reference Guide.
2. After migration, the moved virtual machine may not be accessible over the network for several minutes
due to latencies in the network equipment reconfiguration (for example, when switches need to update
their dynamic VLAN membership tables).
60
Managing Virtual Machines and Containers
Migrating Containers to Virtual Machines
The pmigrate utility allows you to migrate Containers to virtual machines. The source server, i.e.
the server where the Container resides before its migration, can be one of the following:
•
a local server running Parallels Server Bare Metal
•
a remote server running Parallels Server Bare Metal
•
a remote server running Parallels Virtuozzo Containers
Currently, the destination server, i.e. the server where the resulting virtual machine will be created,
can be only a local server with Parallels Server Bare Metal.
The process of migrating a Container to a virtual machine differs depending on whether the server
where the Container resides is running the Parallels Server Bare Metal or Parallels Virtuozzo
Containers software.
Migrating Containers
You can use the pmigrate utility to migrate Containers that reside on both local and remote
servers running Parallels Server Bare Metal. When migrating a Container from a local server, you
only need to specify the Container ID and the name of the resulting virtual machine. For example,
the following command migrates Container 101 to the VM_101 virtual machine on the same
Parallels server:
# pmigrate c 101 v VM_101
Connecting to local agent...
Querying configuration...
Migrating...
Operation progress 100%
Registering VM...
PVC to VM /var/parallels/VM_101.pvm/config.pvs migration succeeded.
The resulting virtual machine will be put to the /var/parallels directory on the destination
server.
If you want to migrate a Container from a remote Parallels server, you should additionally indicate
the source server IP address and the credentials of the root user on this server:
# pmigrate c root:8uhytv4@192.168.12.12/101 v VM_101
Connecting to local agent...
Querying configuration...
Migrating...
Operation progress 100%
Registering VM...
PVC to VM /var/parallels/VM_101.pvm/config.pvs migration succeeded.
This command migrates Container 101 residing on the Parallels server with the IP address of
192.168.12.12 to the VM_101 virtual machine on the local server. If you do not specify the root
credentials on the source server, you will be asked to do so during the command execution.
61
Managing Virtual Machines and Containers
Migrating Containers from a Server with Parallels Virtuozzo Containers
You can use the pmigrate utility to migrate Containers that reside on remote servers running the
following versions of the Parallels Containers software:
•
Parallels Virtuozzo Containers 4.0 for Linux with update TU-4.0.0-464 or higher
•
Parallels Virtuozzo Containers 4.6 for Linux
•
Parallels Virtuozzo Containers 4.5 for Windows
•
Parallels Virtuozzo Containers 4.6 for Windows
Moving a Container from a remote Parallels Containers server to a virtual machine on the local
server with Parallels Server Bare Metal involves completing the following steps:
1
Installing the Parallels agent on the server with Parallels Containers.
2
Running the pmigrate utility on the destination server to migrate the Container.
Installing the Parallels for Containers Agent
First, you must install the Parallels agent on the source Parallels Containers server. During
migration, this agent collects essential information on the Container to be moved and transfers it to
the pmigrate utility on the destination server. To install the Parallels agent, do the following:
1
Log in to the source Parallels Containers server as a user with administrative rights.
2
Copy the Parallels agent installation file to the source server. The installation file is located in the
/usr/share/pmigrate/tools directory on the server with Parallels Server Bare Metal:
• parallels-transporter-for-containers-XXXX.run. Use this file to install the
Parallels agent on servers running Parallels Virtuozzo Containers 4.0 or 4.6 for Linux.
• ParallelsTransporterForContainers-parallels-XXXX.exe. Use this file to
install the Parallels agent on servers running Parallels Virtuozzo Containers 4.5 or 4.6 for
Windows.
3
Execute the installation file on the source server.
4
Follow the instructions of the wizard to install the Parallels agent.
Migrating the Container
Once the Parallels agent is installed, you can use the pmigrate utility to move a Container to a
virtual machine. For example, you can run the following command on the destination server to
migrate Container 101 from the remote server with IP address 192.168.12.12 to the VM_101
virtual machine:
# pmigrate c root:8uhytv4@192.168.12.12/101 v VM_101
Connecting to local agent...
Querying configuration...
Migrating...
Operation progress 100%
62
Managing Virtual Machines and Containers
Registering VM...
PVC to VM /var/parallels/VM_101.pvm/config.pvs migration succeeded.
The resulting virtual machine will be put to the /var/parallels directory on the destination
server. If you do not specify the administrative credentials on the source server (for root on Linux
servers and Administrator on Windows servers), you will be asked to do so during the
command execution.
63
Managing Virtual Machines and Containers
Migrating Physical Computers to Virtual Machines and Containers
You can also use the pmgirate utility to move a stand-alone physical computer to a virtual
machine and Container. The migration process includes copying the whole contents of the physical
computer (i.e. all its files, directories, quota limits, configuration settings, and so on) to a virtual
machine and Container on the Parallels server. After migrating the computer, you will have its exact
copy in a virtual machine and Container including the operating system, the IP addresses assigned,
the amount of available disk space and memory, etc.
Moving a physical computer to a virtual machine and Container involves completing the following
steps:
1
Installing the Parallels agent on the physical computer you want to migrate. This step is required
only if you are migrating the physical computer to a virtual machine.
2
Migrating the physical computer by running the pmigrate utility on the server.
Installing the Agent
If you are planning to migrate a physical computer to a virtual machine, you must first install the
Parallels agent on this computer. This agent collects essential system data on the physical
computer and transfers it to the pmigrate utility on the Parallels server. To install the Parallels
agent, do the following:
1
Make sure that your physical computer meets the necessary requirements for installing the
Parallels agent. See Requirements for Migrating to Virtual Machines (p. 69) for details.
2
Log in to the physical computer as a user with administrative rights.
3
Copy the Parallels agent installation file to the physical computer. The installation file is located
in the /usr/share/pmigrate/tools directory on the Parallels server:
• parallels-transporter-agent-XXXX.run. Use this file to install the Parallels agent
on computers running a Linux operating system.
• ParallelsTransporterAgent-parallels-XXXX.exe. Use this file to install the
Parallels agent on computers running a Windows operating system.
4
Execute the installation file on the physical computer.
5
Follow the instructions of the wizard to install the Parallels agent.
6
Restart the source computer to complete the installation.
Note: The Parallels agent is automatically launched after the restart, so you do not need to start it
manually.
64
Managing Virtual Machines and Containers
Migrating the Server
Once the physical computer is up and running, you can migrate to a virtual machine and Container
on the Parallels server. For example, you can move a physical computer to a virtual machine by
running the following command on the destination server:
# pmigrate h root:1qsde34rt@192.168.1.130 v MyVM
where
•
h denotes that you are migrating a physical computer.
•
root:1qsde34rt@192.168.1.130 is the IP address and credentials of the physical
computer to be migrated.
You can omit the credentials in the command above. In this case you will be asked to provide
them during the command execution.
•
v indicates that the physical computer is to be moved to a virtual machine.
•
MyVM is the name of the resulting virtual machine on the Parallels server.
Once the command is complete, you will find the resulting virtual machine in the
/var/parallels directory on the Parallels server.
If you want to migrate the same physical computer to a Container, just specify c instead of v and
the ID of the resulting Container (e.g. 101) instead of MyVM. For example, the following command
will migrate the physical computer to Container 101:
# pmigrate h root:1qsde34rt@192.168.1.130 c 101
Notes:
1. Migrating physical computers running a Windows operating system to Containers is not supported.
2. Migrating physical computers with system volumes formatted with ReiserFS is not supported.
65
Managing Virtual Machines and Containers
Requirements for Migrating to Containers
To avoid delays and problems when migrating a physical server to a Container, make sure that the
following requirements are fulfilled:
Migrating to Containers on Linux servers:
•
The Linux distribution installed on the physical server is supported by Parallels Server Bare
Metal. To find out if your Linux distribution can be recognized by Parallels Server Bare Metal,
you can check the /etc/vz/conf/dists directory on the Parallels server and look for the
configuration file of your Linux distribution. It should have the name of
Linux_Distribution_Name-version.conf where Linux_Distribution_Name and
version denote the name of the Linux distribution and its version, respectively (e.g. redhat5.conf). If there is no corresponding distribution in the directory, you can do one of the
following:
• Create a new distribution configuration file and place it to the /etc/vz/conf/dists
directory on the Parallels server. Detailed information on how to create new configuration
files is provided in the Creating Configuration Files for New Linux Distribution section (p.
234).
• Start the migration process without having the right configuration file for your Linux
distribution. In this case the unknown.conf distribution configuration file from the
/etc/vz/conf/dists directory will be used for tuning the Container after the physical
server migration. However, using the unknown.conf configuration file means that you will
not be able to use standard Parallels Server Bare Metal utilities (e.g. pctl) for performing
the main operations on the created Container (such as setting the Container IP address or
configuring the DNS parameters) and have to manually complete these tasks from inside the
Container.
•
ssh is installed on both the physical server and the Parallels. ssh is used to provide secure
encrypted and authenticated communication for both physical servers. You can check if the
ssh package is already installed on the server by executing the ssh -V command.
•
rsync is installed on the physical server. rsync is used to copy the physical server contents
to the Container. If the physical server rsync happens to be incompatible with the Parallels
server, use the statically linked rsync from the /usr/local/share/vzlinmigrate
directory on the physical server as well.
Migrating to Containers on Windows servers:
Migrating physical computers running a Windows operating system to Containers is not supported.
66
Managing Virtual Machines and Containers
Migration Restrictions for Containers
Listed below are the limitations you should take into account when deciding on the migration
process.
Migrating to Containers on Linux servers:
•
During the migration, all the filesystems available on your physical server are joined to one
filesystem inside the Container - VZFS (Virtuozzo File System). Detailed information on VZFS is
provided in the Virtuozzo File System subsection (p. 18).
•
If there are several IP addresses assigned to the physical server, all these IP addresses will be
reassigned to one and the same device on the Parallels server - venet0. This virtual network
adapter is used to connect all the Containers on the given Parallels server among themselves
and with the server. After the migration, you can create additional virtual network adapters
inside the Container and decide what IP address to be assigned to what network adapter. For
detailed information on how to create and manage Container virtual network adapters, turn to
Managing Adapters in Containers (p. 179).
•
During the migration process, you may specify only one partition on the physical server which
will be migrated to the Container together with all quotas imposed on it. All the other partitions
of the server will be copied without keeping their quota limits. Moreover, the quota limits
imposed on the selected partition will be applied to the entire Container after the server
migration.
•
While migrating your physical server running a Linux operating system with the securityenhanced (SE) Linux kernel, keep in mind that the SE Linux kernel is currently not supported by
Parallels Server Bare Metal. Therefore, the Container where the server running the SE Linux
distribution has been migrated will not support the SE security features.
•
If any of your files and/or directories on the physical server have extended attributes associated
with them, these attributes will be lost after the server migration.
•
Raw devices on the physical server cannot and will not be migrated to the Container on the
Parallels server.
•
If you are running an application which is bound to the physical server MAC address, you will
not be able to run this application inside the Container after the server migration. In this case,
you can do one of the following:
• If you are running a licensed application, you should obtain a new license and install the
application inside the Container anew.
• If you are running a non-licensed application, you can try to reconfigure the application and
to make it work without being bound to any MAC address.
•
If the migration process fails on the step of transferring files and directories from the physical
server to the Container by means of rsync, the /vz/private/CT_ID directory on the
Parallels server will contain all the copied files and directories and may occupy a great amount
of disk space. You can keep the directory, which will greatly speed up the repeated migration
procedure, or manually remove the directory by using the rm utility.
•
Migrating physical computers with system volumes formatted with ReiserFS is not supported.
67
Managing Virtual Machines and Containers
Migrating to Containers on Windows servers:
Migrating physical computers running a Windows operating system to Containers is not supported.
68
Managing Virtual Machines and Containers
Requirements for Migrating to Virtual Machines
Any physical computer that you plan to migrate to a virtual machine must have the Parallels agent
installed. The agent can be installed on computers meeting the following requirements.
Note: Migrating physical computers with system volumes formatted with ReiserFS is not supported.
Hardware Requirements
•
700 (or higher) MHz x86 or x64 processor (Intel or AMD).
•
256 MB or more RAM.
•
50 MB of hard disk space for installing the Parallels agent package.
•
Ethernet or WiFi network adapter.
Software Requirements
For software requirements, see the table in General Migration Requirements (p. 55).
Additional Requirements for Migrating Servers with Parallels Server Bare Metal
If you plan to migrate a server running the Parallels Server Bare Metal software, you should first
make sure that the snapapi26 and snumbd26 modules are not loaded on the server. You can
use the following command to check this:
# lsmod | grep snapapi26
# lsmod | grep snumbd26
If any of these modules are loaded, unload them by running the rmmod command. If the rmmod
command fails to unload any of these modules, do the following:
1
Stop the Parallels Server Bare Metal service:
# /etc/init.d/vz stop
2
Unload the modules:
# rmmod snapapi26
# rmmod snumbd26
3
Start the Parallels Server Bare Metal service again:
# /etc/init.d/vz start
Once the modules are unloaded, proceed with migrating the server.
Notes:
1. Migrating Windows dynamic volumes and Linux logical volumes (LVM) is not supported.
2. You may also try to migrate servers with unsupported file systems. However, in this case all disk
sectors are copied successively, and you may experience problems with using the resulting virtual
machine.
69
Managing Virtual Machines and Containers
Migrating Virtual Machines to Containers
The process of migrating a virtual machine to a Container on the Parallels server is the same as
migrating a physical computer to a Container. For example, you can execute the following
command to move a virtual machine with the IP address of 192.168.1.130 to Container 101 on
your Parallels server:
# pmigrate h root:1qsde34rt@192.168.1.130 c 101
You can omit the virtual machine credentials in the command above. In this case you will be asked
to provide them during the command execution.
Notes:
1. For more information on migrating physical computers to Containers, see Migrating Physical
Computers to Virtual Machines and Containers (p. 64).
2. The requirements a virtual machine must meet are the same as for migrating physical computers; they
are described in Requirements for Migrating to Containers (p. 66).
70
Managing Virtual Machines and Containers
Migrating Xen Virtual Machines
You can use the pmgirate utility to migrate Xen virtual machines to virtual machines on the
Parallels server. Moving a Xen virtual machine to a Parallels virtual machine involves completing the
following steps:
1
Installing the Parallels agent on the physical server hosting the Xen virtual machine.
2
Migrating the Xen virtual machine by running the pmigrate utility on the server.
Both steps are described below in detail.
Installing the Agent
Before you start migrating a Xen virtual machine, you need first to install the Parallels agent on the
Xen server where the virtual machine is residing. To install the Parallels agent, do the following:
1
Log in to the Xen server as a user with administrative rights.
2
Copy the Parallels agent installation file to the Xen server. The installation file is located in the
/usr/share/pmigrate/tools directory on the Parallels server and has the name
parallels-transporter-agent-parallels-en_US-XXXX.run.
3
Execute the copied file, and follow the instructions to install the Parallels agent.
4
Start the Parallels agent:
# parallels-transporter-agent -c
Migrating the Xen virtual machine
Once the Parallels agent is running on the Xen server, you can migrate the Xen virtual machine. Let
us assume the following:
•
You want to migrate the XenVM virtual machine from the Xen server to the MigratedVM virtual
machine on the Parallels server.
•
root:1qsde34rt@192.168.1.130 is the IP address and credentials of the Xen server
where the MigratedVM virtual machine resides.
To migrate the XenVM virtual machine, you can run the following command:
# pmigrate x root:1qsde34rt@192.168.1.130/XenVM v MigratedVM
In this command, x denotes that you are migrating a Xen virtual machine, and v indicates that the
Xen virtual machine is to be moved to a Parallels virtual machine. If you omit the credentials in the
command above, you will be asked to provide them during the command execution. Once the
migration is complete, you can find the resulting virtual machine in the /var/parallels directory
on the Parallels server.
Note: You are recommended to check the settings of the migrated virtual machine (for example, memory
and network settings) and, if necessary, configure them to meet your needs.
71
Managing Virtual Machines and Containers
Troubleshooting the migration of paravirtualized Xen virtual machines
When migrating a paravirtualized Xen virtual machine, pmigrate first copies the whole of the
virtual machine to the Parallels server and then replaces the paravirtualized kernel of the copied
machine with a normal kernel from the corresponding Linux distribution. If it cannot replace the
kernel, pmigrate displays an error but does not delete the virtual machine from the Parallels
server. In this case, you can do the following:
•
Remove the copied virtual machine from the Parallels server and try to migrate the virtual
machine again.
•
Configure the copied virtual machine on the Parallels server manually.
If you choose the second way, do the following:
1
Boot into the virtual machine in rescue mode using an ISO image of the Linux OS
corresponding to the OS installed in the virtual machine.
2
Detect where on the disk the root partition is located, and mount it.
3
Detect all other partitions on the disk (/boot, /usr, and so on), and mount them to the
corresponding directories on the root partition; also mount the /proc file system.
4
Install a normal Linux kernel (for example, from the ISO image you used to boot into the virtual
machine). The normal kernel must be of the same architecture as the paravirtualized Xen kernel.
5
Create the initrd image for the normal kernel if you have not already done so when installing
the kernel.
6
Configure the bootloader to load the normal kernel if you have not already done so when
installing the kernel.
7
Configure the /etc/inittab file to start getty and tty1-tty6.
8
Unmount the partitions.
9
Restart the virtual machine, and boot into the normal kernel.
Performing Container-Specific Operations
This section provides the description of operations specific for your Containers.
72
Managing Virtual Machines and Containers
Setting Name for Container
You can assign an arbitrary name to a Container and use it, along with the Container ID, to perform
Container-related operations. For example, you can start or stop a Container by specifying the
Container name instead of its ID.
You can assign names to Containers using the --name option of the pctl set command. For
example, to set the computer1 name for Container 101, run this command:
# pctl set 101 --name computer1 --save
Name computer1 assigned
Saved parameters for Container 101
You can also set a name for Container 101 by editing its configuration file:
1
Open the configuration file of Container 101 (/etc/vz/conf/101.conf) for editing, and add
the following string to the file:
NAME="computer1"
2
In the /etc/vz/names directory on the server, create a symbolic link with the name of
computer1 pointing to the Container configuration file. For example:
# ln --symbolic /etc/vz/conf/101.conf /etc/vz/names/computer1
When specifying names for Containers, keep in mind the following:
•
Names may contain the following symbols: a-z, A-Z, 0-9, underscores (_), dashes (-),
spaces, the symbols from the ASCII character table with their code in the 128-255 range, and
all the national alphabets included in the Unicode code space.
•
Container names cannot consist of digits only. Otherwise, there would be no way to distinguish
them from Container IDs.
•
If it contains one or more spaces, the Container name must be put in single or double quotes.
Once you assign the computer1 name to Container 101, you can start using it instead of ID 101
to perform Container-related operations. For example:
•
You can stop Container 101 with the following command:
# pctl stop computer1
Stopping Container ...
Container was stopped
Container is unmounted
•
You can start Container 101 anew by running the following command:
# pctl start computer1
Starting Container ...
...
You can find out what name is assigned to Container 101 in one of the following ways:
•
Using the vzlist utility:
# vzlist -o name 101
NAME
computer1
73
Managing Virtual Machines and Containers
•
Checking the NAME parameter in the Container configuration file (/etc/vz/conf/101.conf):
# grep NAME /etc/vz/conf/101.conf
NAME="computer1"
•
Checking the NAME parameter in the /etc/vz/names/computer1 file which is a symlink to
the Container configuration file:
# grep NAME /etc/vz/names/computer1
NAME="computer1"
Moving Container Within the Parallels Server
The vzmlocal utility allows you to move Containers within your server. Moving a Container within
one and the same server consists in changing the Container ID and its private area and root paths.
You can use vzmlocal to change the ID of the corresponding Container only or to additionally
modify its private area and root path.
Let us assume that you want to change the ID of your Container from 101 to 111 and modify its
private area and root paths from /vz/private/101 to /vz/private/my_dir and from
/vz/root/101 to /vz/root/ct111, respectively. To do this, execute the following command
on the server:
# vzmlocal 101:111:/vz/private/my_dir:/vz/root/ct111
Moving/copying Container#101 -> Container#111,
[/vz/private/my_dir], [/vz/root/ct111] ...
...
Successfully completed
To check if Container 101 has been successfully moved to Container 111, you can use the
following commands:
# vzlist -a
CTID
NPROC STATUS IP_ADDR
1
43 running 10.0.10.1
111
- stopped 10.0.10.101
# ls /vz/private
1 my_dir
# ls /vz/root
1 ct111
HOSTNAME
localhost
myContainer
The commands output shows that the ID of Container 101 has been changed to 111, its private
area is now located in the /vz/private/my_dir directory on the server, and the path to its root
directory is /vz/root/ct111.
Notes:
1. You can use the vzmlocal utility to move several Containers simultaneously.
2. You can run the vzmlocal utility on both running and stopped Containers.
74
Managing Virtual Machines and Containers
Disabling Container
There may appear situations when you need to forbid Container owners to use their Containers.
For example, it may happen if the Container owner uses it for unallowed purposes: intruding into
computers of other users, participating in DoS attacks, and so on.
In such cases, you can disable a Container, thus making it impossible to start the Container once it
was stopped. For example, you can execute the following command to disable Container 101:
# pctl set 101 --disabled yes
Once Container 101 is stopped, the user will not be able to start it again until you enable the
Container again:
# pctl set 101 --disabled no
You can also use the --force option to start a disabled Container. For example:
# pctl start 101
Container start disabled
# pctl start 101 --force
Starting Container...
Container is mounted
Adding port redirection to Container(1): 4643 8443
Adding IP address(es): 10.144.144.101
Hostname for Container set: Container_101
Container start in progress...
75
Managing Virtual Machines and Containers
Reinstalling Container
Reinstalling a Container is used if a Container administrator has inadvertently modified, replaced, or
deleted any file that is part of an application or OS template, which has brought about the
Container malfunction. You can reinstall the Container in the two following ways:
1
The pctl recover command restores the original VZFS symlinks of the Container private
area to the OS and/or application template(s) as they were at the time when the Container was
created and/or when the application template(s) were added to the Container. This command
does not deal with any user files on the Container:
# pctl recover 101
Optimizing Container private area...
vzquota : (warning) Quota is running for id 101 already
Setting quota ...
Container is mounted
Container is unmounted
Recover OS template: redhat-el5-x86
Creating Container private area (redhat-el5-x86)
...
Recovering Container completed successfully
2
The pctl reinstall command creates a new private area for the problem Container from
scratch using its configuration files and its OS and application templates. Thus, a clean working
copy of the Container is created:
# pctl reinstall 101
Optimizing Container private area...
Calculating Container disk usage...
Creating Container private area (redhat-el5-x86)
Starting Container ...
Initializing quota...
Container is mounted
Container start in progress...
Calculating Container disk usage...
Copying Container credentials...
Stopping Container ...
Container was stopped
Container is unmounted
Old Container file system has been moved to /old
Initializing quota...
Container reinstallation completed successfully
Note: If any of the Container application templates cannot be added to the Container in a normal way,
the reinstallation process will fail. This may happen, for example, if an application template was added to
the Container using the --force option of the vzpkgadd or vzpkg install command (for more
information on these commands, see the Parallels Command Line Reference Guide).
In order to retain the personal data inside the old Container, the utility also copies the contents of
the old private area to the /old directory of the new private area (unless the --skipbackup
option is given). The personal data can then be copied to the corresponding directories of the new
private area and the /old directory eventually deleted:
# pctl start 101
Starting Container ...
Container is mounted
76
Managing Virtual Machines and Containers
Setting devperms 20002 dev 0x7d00
Adding port redirection to Container(1): 4643 8443
Adding IP address(es) to pool:
Adding IP address(es): 10.14.14.101
Hostname for Container set: localhost.localdomain
Container start in progress...
# pctl exec 101 ls /
bin
boot
dev
[...other directories...]
old
[...other directories...]
tmp
usr
var
Both the pctl recover and pctl reinstall commands retain the users' credentials base,
unless the --resetpwdb option is specified.
77
Managing Virtual Machines and Containers
Customizing Container Reinstallation
The default reinstallation, as performed by the pctl reinstall command, creates a new private
area for the broken Container as if it were created by the pctl create command and copies the
private area of the broken Container to the /old directory in the new private area so that no file is
lost. There is also a possibility of deleting the old private area altogether without copying or
mounting it inside the new private area, which is done by means of the --skipbackup option.
This way of reinstalling corrupted Containers might in certain cases not correspond exactly to your
particular needs. It happens when you are accustomed to creating new Containers in some other
way than just using the pctl create command. For example, you may install additional software
licenses into new Containers, or anything else. In this case you would naturally like to perform
reinstallation in such a way so that the broken Container is reverted to its original state as
determined by you, and not by the default behavior of the pctl create command.
To customize reinstallation, you should write your own scripts determining what should be done
with the Container when it is being reinstalled, and what should be configured inside the Container
after it has been reinstalled. These scripts should be named vps.reinstall and
vps.configure, respectively, and should be located in the /etc/vz/conf directory on the
server. To facilitate your task of creating customized scripts, the Containers software is shipped
with sample scripts that you may use as the basis of your own scripts.
When the pctl reinstall <CT_ID> command is called, it searches for the vps.reinstall
and vps.configure scripts and launches them consecutively. When the vps.reinstall
script is launched, the following parameters are passed to it:
--veid
The ID of the Container.
--ve_private_tmp
The path to the Container temporary private area. This path designates where
a new private area is temporarily created for the Container. If the script runs
successfully, this private area is mounted to the path of the original private area
after the script has finished.
--ve_private
The path to the Container original private area.
You may use these parameters within your vps.reinstall script.
If the vps.reinstall script finishes successfully, the Container is started, and the
vps.configure script is called. At this moment the old private area is mounted to the /old
directory inside the new one irrespective of the --skipbackup option. This is done in order to let
you use the necessary files from the old private area in your script, which is to be run inside the
running Container. For example, you might want to copy some files from there to regular Container
directories.
After the vps.configure script finishes, the old private area is either dismounted and deleted or
remains mounted depending on whether the --skipbackup option was provided.
If you do not want to run these reinstallation scripts and want to stick to the default pctl
reinstall behavior, you may do either of the following:
78
Managing Virtual Machines and Containers
1
Remove the vps.reinstall and vps.configure scripts from the /etc/vz/conf
directory, or at least rename them;
2
Modify the last line of the vps.reinstall script so that it would read
exit 128
instead of
exit 0
The 128 exit code tells the utility not to run the scripts and to reinstall the Container with the default
behavior.
79
Managing Virtual Machines and Containers
Performing Virtual Machine-Specific Operations
This section focused on operations specific for your virtual machines.
Viewing Detailed Information About Virtual Machines
To view detailed information about a virtual machine, you can use the pctl list -i command.
For example, the following command lists all information about the MyVM virtual machine:
# pctl list -i MyVM
ID: {5c1fb1bb-4364-4b42-86b2-c584bdd2223b}
EnvID: 2075205468
Name: MyVM
Description:
State: running
OS: win-7
Uptime: 21:49:55 (since 2011-05-19 12:25:19)
Home: /var/parallels/MyVM.pvm/
Owner: root@.
Effective owner: owner
GuestTools: state=installed
Autostart: off
Autostop: shutdown
Boot order: hdd0 cdrom0 fdd0
Remote display: mode=off port=6500 address=0.0.0.0
Remote display state: stopped
Hardware:
cpu 1 VT-x accl=high mode=32 ioprio=4 iolimit=0
memory 1024Mb
video 32Mb
memory_quota auto
fdd0 (+) real='/dev/fd0' state=disconnected
hdd0 (+) sata:0 image='/var/parallels/MyVM.pvm/MyVM-0.hdd' 65536Mb
cdrom0 (+) sata:1 real='D: ' state=disconnected
parallel0 (+) real='/dev/lp0'
usb (+)
net0 (+) dev='vme7bb11f5c.0' network='Bridged' mac=001C427B68E3 card=e1000
Host Shared Folders: (+)
SmartMount: (-)
VirtualUsbMouse: state=disabled
Encrypted: no
Offline management: (-)
The following table describes the main options displayed by pctl list -i.
Option Name
Description
ID
Virtual machine identifier. Usually, you use this ID, along with the
virtual machine name, when performing an operation on the virtual
machine.
EnvID
Kernel virtual machine identifier. This is the ID the kernel on the
physical server uses to refer to a virtual machine when displaying
some information on this virtual machine.
Name
Virtual machine name.
80
Managing Virtual Machines and Containers
Description
Virtual machine description.
State
Virtual machine state.
OS
Guest operating system installed in a virtual machine.
Uptime
Time a virtual machine has been up and running from its last start.
Home
Directory storing virtual machine files.
Guest Tools
Shows whether Parallels Tools are installed in a virtual machine.
Autostart
Shows whether a virtual machine is automatically started when you
turn on the physical server.
Boot order
Order in which the virtual machine devices are checked for an
operating system.
Hardware
Devices available in a virtual machine.
Pausing a Virtual Machine
Pausing a running virtual machine releases the resources, such as RAM and CPU, currently used
by this virtual machine. The released resources can then be used by the Parallels server or other
running virtual machines and Containers.
To pause a virtual machine, you can use the pctl pause command. For example, the following
command pauses the My_VM virtual machine:
# pctl pause My_VM
Pause the VM...
The VM has been successfully paused.
You can check that the virtual machine has been successfully paused by using the pctl list a command:
# pctl list -a
STATUS
IP_ADDR
running 10.10.10.101
paused
10.10.10.201
NAME
101
My_VM
The command output shows that the My_VM virtual machine is paused at the moment. To continue
running this virtual machine, execute this command:
# pctl start My_VM
Starting the VM...
The VM has been successfully started.
81
Managing Virtual Machines and Containers
Managing Snapshots
In Parallels Server Bare Metal, you can save the current state of a virtual machine by creating a
snapshot. You can then continue working in your virtual machine and return to the saved state any
time you wish. For example, you can make use of snapshots in the following cases:
•
You are going to configure an application with a lot of settings. In this case, you may first wish
to play with settings before applying them to your application. So, you create a snapshot before
starting to experiment with the application settings.
•
You are involved in a large development project. In this case, you may wish to mark milestones
in the development process by creating a snapshot after each milestone. If anything goes
wrong, you can easily revert to the previous milestone and start the development anew.
In Parallels Server Bare Metal, you can manage snapshots as follows:
•
create a new snapshot of a virtual machine
•
list the existing snapshots of a particular virtual machine
•
revert to a snapshot
•
remove a snapshot
All these operations are described in the following subsections in detail.
82
Managing Virtual Machines and Containers
Creating a Snapshot
To create a snapshot of a virtual machine in Parallels Server Bare Metal, you can use the pctl
snapshot command. For example, you can execute the following command to create a snapshot
of the MyVM virtual machine:
# pctl snapshot MyVM
Creating the snapshot...
The snapshot with ID {12w32198-3e30-936e-a0bbc104bd20} has been successfully created.
A newly created snapshot is saved to the /vz/VM_Name.pvm/Snapshots/Snapshot_ID.pvs
file where VM_Name denotes the name of the corresponding virtual machine and Snapshot_ID is
a random ID assigned to the snapshot. In the command above, the snapshot is assigned the ID of
{12w32198-3e30-936e-a0bbc104bd20} and saved to the
/vz/MyVM/Snapshots/{12w32198-3e30-936e-a0bbc104bd20}.pvs file.
# ls /vz/MyVM.pvm/Snapshots/
{063615fa-f2a0-4c14-92d4-4c935df15840}.pvc
The ID assigned to the snapshot can be used to manage this snapshot (e.g. get detailed
information on the snapshot or delete it) .
When creating a snapshot, you can also set a name for it and provide its description:
# pctl snapshot MyVM -n Clean_System -d "This snapshot was created right after
installing the Windows XP operating system"
Creating the snapshot...
The snapshot with ID {0i8798uy-1eo0-786d-nn9ic106b9ik} has been successfully created.
You can then view the set name and description in the /vz/MyVM/Snapshots.xml file or in
Parallels Management Console.
When working with snapshots, keep in mind the following:
•
If a virtual machine name contains spaces, use quotation marks to specify the name in pctl
commands (e.g. "Windows XP").
•
Before creating a snapshot, it is recommended that you complete all operations of installing,
downloading, or writing to external devices. You should also complete or cancel any
transactions performed via the virtual machine in external databases.
Creating Branches
The branches are created when you do the following:
1
Create several sequential snapshots.
2
Revert to an intermediate snapshot.
3
Make some changes to the virtual machine.
4
Save the virtual machine state by creating a new snapshot.
In this case, the newly created snapshot will start a new branch using the intermediate snapshot
from Step 2 as the baseline.
83
Managing Virtual Machines and Containers
Listing Snapshots
To list all snapshots of a particular virtual machine, use the pctl snapshot-list command:
# pctl snapshot-list MyVM
PARENT_SNAPSHOT_ID
SNAPSHOT_ID
{989f3415-3e30-4494-936e-a0bbc104bd20}
{989f3415-3e30-4494-936e-a0bbc104bd20} *{063615fa-f2a0-4c14-92d4-4c935df15840}
This command shows that currently two snapshots exist for the MyVM virtual machine. The
snapshot with ID {063615fa-f2a0-4c14-92d4-4c935df15840} is based on the snapshot
with ID {989f3415-3e30-4494-936e-a0bbc104bd20}, i.e. the latter acts as the parent for
the snapshot with ID {063615fa-f2a0-4c14-92d4-4c935df15840}. The * sign before
{063615fa-f2a0-4c14-92d4-4c935df15840} denotes that this is the current snapshot for
the given virtual machine.
You can also view the relationship between snapshots by specifying the -t option:
# pctl snapshot-list MyVM -t
_{989f3415-3e30-4494-936e-a0bbc104bd20}_{063615fa-f2a0-4c14-92d4-4c935df15840}
\*{712305b0-3742-4ecc-9ef1-9f1e345d0ab8}
The command output shows you that currently 2 branches exist for the MyVM virtual machine. The
snapshot with ID {989f3415-3e30-4494-936e-a0bbc104bd20} is the baseline used as a
starting point for these branches.
You can get detailed information on a particular snapshot using the -i option and specifying the
snapshot ID:
# pctl snapshot-list MyVM -i {063615fa-f2a0-4c14-92d4-4c935df15840}
ID: {063615fa-f2a0-4c14-92d4-4c935df15840}
Name: Clean_System
Date: 2009-07-22 22:39:06
Current: yes
State: power_off
Description: <![CDATA[This snapshot was created right after installing Windows XP
operating system]]>
The pctl snapshot-list command displays the following information about snapshots:
Field
Description
ID
The ID assigned to the snapshot.
Name
The name assigned to the snapshot.
Date
The date and time when the snapshot was created.
Current
Denotes whether this is the current snapshot of the virtual machine.
State
The state the virtual machine was in at the time you took the
snapshot.
Description
The description set for the snapshot.
84
Managing Virtual Machines and Containers
Reverting to a Snapshot
You can use the pctl snapshot-switch command to revert to a snapshot. When you revert to
a snapshot, the current state of the virtual machine is discarded, and all changes made to the
system since the previous snapshot are lost. So, before returning to a specific snapshot, you may
first wish to save these states by creating a new snapshot. Refer to the Creating a Snapshot
subsection (p. 83) for information on how you can do it.
The pctl snapshot-switch command requires the virtual machine name and the snapshot ID
to be specified as arguments:
pctl snapshot-switch "Windows XP" --id {cedbc4eb-dee7-42e2-9674-89d1d7331a2d}
Switch to the snapshot...
The VM has been successfully switched.
This command restores the snapshot with ID {cedbc4eb-dee7-42e2-9674-89d1d7331a2d}
for the Windows XP virtual machine.
Deleting a Snapshot
In Parallels Server Bare Metal, you can use the pctl snapshot-delete command to delete
those snapshots that you do not need any more. Assuming that you want to delete the snapshot
with ID {903c12ea-f6e6-437a-a2f0-a1d02eed4f7e} for the MyVM virtual machine, you can
run this command:
# pctl snapshot-delete MyVM --id {903c12ea-f6e6-437a-a2f0-a1d02eed4f7e}
Deleting the snapshot...
The snapshot has been successfully deleted.
When you delete a parent snapshot, its children are not deleted, and the information the parent
snapshot contains is merged into them.
For example, the following session demonstrates the process of deleting the snapshot with ID
{903c12ea-f6e6-437a-a2f0-a1d02eed4f7e} acting as a parent for another snapshot:
# pctl snapshot-list MyVM
PARENT_SNAPSHOT_ID
SNAPSHOT_ID
{989f3415-3e30-4494-936e-a0bbc104bd20}
{989f3415-3e30-4494-936e-a0bbc104bd20} {063615fa-f2a0-4c14-92d4-4c935df15840}
{063615fa-f2a0-4c14-92d4-4c935df15840} *{58c9941e-f232-4273-892a-82e836536889}
# pctl snapshot-delete MyVM --id {903c12ea-f6e6-437a-a2f0-a1d02eed4f7e}
Deleting the snapshot...
The snapshot has been successfully deleted.
# pctl snapshot-list MyVM
PARENT_SNAPSHOT_ID
SNAPSHOT_ID
{063615fa-f2a0-4c14-92d4-4c935df15840}
{063615fa-f2a0-4c14-92d4-4c935df15840} *{58c9941e-f232-4273-892a-82e836536889}
85
Managing Virtual Machines and Containers
Managing Templates
A template in Parallels Server Bare Metal is a pre-configured virtual machine that can be easily and
quickly deployed into a fully functional virtual machine. Like any normal virtual machine, a template
contains hardware (virtual disks, peripheral devices) and the operating system. It can also have
additional software installed. In fact, the only main difference between a virtual machine and a
template is that the latter cannot be started.
In Parallels Server Bare Metal, you can perform the following operations on templates:
•
create a new template
•
list the existing templates
•
create a virtual machine from a template
These operations are described in the following subsections in detail.
Create a Template
In Parallels Server Bare Metal, you can create a virtual machine template using the pctl clone
utility. Making a template may prove useful if you need to create several virtual machines with the
same configuration. In this case, your steps can be as follows:
1
You create a virtual machine with the required configuration.
2
You make a template on the basis of the created virtual machine.
3
You use the template to create as many virtual machines as necessary.
Let us assume that you want to create a template of the My_VM virtual machine. To do this, you
can run the following command:
# pctl clone My_VM --name template1 --template
Clone the My_VM VM to VM template template1...
Operation progress 98%
The VM has been successfully cloned.
This command clones the My_VM virtual machine and saves it as the template1 template. After
the template has been successfully created, you can use it for creating new virtual machines.
86
Managing Virtual Machines and Containers
Listing Templates
Sometimes, you may need to get an overview of the virtual machine templates available on your
Parallels server. For example, this may be necessary if you plan to create a virtual machine from a
specific template, but do not remember its exact name. In this case, you can use the pctl list
command to list all templates on the Parallels server and find the necessary one:
# pctl list -t
{4ad11c28-9f0e-4086-84ea-9c0487644026}
{64bd8fea-6047-45bb-a144-7d4bba49c849}
{6d3c9d6f-921a-484d-9772-bc7096f68df1}
win-2003
rhel
win-2003
template1
template3
template2
In this example, 3 virtual machine templates exist on the Parallels server. The information on these
templates is presented in the form of a table with the following columns (from left to right): the
template ID, the operating system contained in the template, and the template name.
Deploying a Template
Though a template has the same components as a virtual machine (hardware, software, etc.), it
cannot be started. To run a template as a virtual machine, you need first to convert the template.
By converting a template, you create a virtual machine with the configuration identical to that of the
template.
To convert a template into a virtual machine, use the --ostemplate option of the pctl
create command. For example, to convert the template1 template to a virtual machine with the
Converted_VM name, you can run this command:
# pctl create Converted_VM --ostemplate template1
Creating the VM on the basis of the template1 template...
Clone the template1 VM to the VM Converted_VM...
Operation progress 99%
The VM has been successfully cloned.
To check that the Converted_VM virtual machine has been successfully created, use the pctl
list -a command:
# pctl list -a
STATUS
IP_ADDR
running
10.12.12.101
stopped
10.12.12.34
running
10.30.17.149
NAME
111
Converted_VM
Windows XP
The template itself is left intact and can be used for creating other virtual machines:
# pctl list -t
{4ad11c28-9f0e-4086-84ea-9c0487644026}
{64bd8fea-6047-45bb-a144-7d4bba49c849}
win-2003
rhel
template1
template2
87
Managing Virtual Machines and Containers
Managing Virtual Machine Disks
In Parallels Server Bare Metal, you can manage virtual machine disks as follows:
•
change the type of your virtual disks
•
increase the capacity of your virtual disks
•
reduce the capacity of your virtual disks
•
reduce the size occupied by your virtual disks on the physical hard drive
All these operations are described in the following subsections in detail.
Changing the Disk Type
A virtual disk can be one of the two types:
•
plain. A plain virtual hard disk has a fixed size from the moment of its creation.
•
expanding. An expanding virtual hard disk is small initially. Its size grows as you add
applications and data to it.
A new virtual machine is created with an expanding virtual disk. However, you can change the type
of the virtual disk using either the pctl or prl_disk_tool utility. Let us assume that the current
type of the hdd0 virtual disk in the MyVM virtual machine is expanding and you want to change it to
plain. To do this, you can execute one of the following commands:
# pctl set MyVM --device-set hdd0 --type plain
or
# prl_disk_tool convert --hdd /vz/vmprivate/MyVM/harddisk.hdd --plain
The main difference between these two commands is that pctl requires for its execution the disk
name as it is shown by the pctl list --info command (hdd0) while prl_disk_tool needs
the full path to the virtual disk drive (/vz/vmprivate/MyVM/harddisk.hdd).
To change the disk type back to expanding, run one of the following commands:
# pctl set MyVM --device-set hdd0 --type expand
or
# prl_disk_tool convert --hdd /vz/vmprivate/MyVM/harddisk.hdd --expanding
88
Managing Virtual Machines and Containers
Increasing the Disk Capacity
If you find that the capacity of your virtual machine's hard disk does not fit your needs anymore,
you can increase it using the prl_disk_tool utility. For example, you can execute the following
command to set the capacity for the MyVM-0.hdd disk to 80 GB:
# prl_disk_tool resize --size 80000 --hdd /vz/MyVM.pvm/MyVM-0.hdd/
Operation progress 100%
This command adds additional disk space as unallocated space. You can use standard means
(e.g. the Disk Management tool in Windows-based virtual machines) to allocate this space by
creating a new partition or expanding an existing on.
At the same time, you can use the --resize_partition option to automatically add additional
space to the last partition on the virtual disk:
# prl_disk_tool resize --size 80000 --hdd /vz/MyVM.pvm/MyVM-0.hdd/ --resize_partition
Operation progress 100%
When increasing the disk capacity, keep in mind the following:
•
You cannot increase the capacity of a virtual disk if the virtual machine using this disk is
running.
•
The virtual machine using the virtual disk you want to configure must not have any snapshots.
Otherwise, the operation will fail:
# prl_disk_tool resize --size 68000 --hdd /vz/MyVM.pvm/MyVM-0.hdd/
This disk has one or more snapshots and cannot be resized.
You need to delete snapshots using the pctl tool before resizing the disk.
In this case, you should delete all existing snapshots and run the command again. To learn how
to delete virtual machine snapshots, refer to Deleting a Snapshot (p. 85).
•
The capacity of an expanding virtual disk shown from inside the virtual machine and the size the
virtual disk occupies on the server's physical disk may differ.
89
Managing Virtual Machines and Containers
Reducing the Disk Capacity
Parallels Server Bare Metal provides a possibility to reduce the size of an expanding virtual disk by
setting the limit the disk cannot exceed. In general, the process of reducing a virtual disk includes
these steps:
1
Finding out the minimum capacity to which the disk can be reduced.
2
Running the prl_dsk_tool resize command to reduce the disk.
Checking the Minimum Disk Capacity
Before reducing a virtual disk, you may wish to see the minimum capacity to which it can be
reduced. To do this, use the prl_disk_tool resize --info command. For example, you
can run the following command to get detailed information on the MyVM-0.hdd disk:
# prl_disk_tool resize --info --hdd /vz/MyVM.pvm/MyVM-0.hdd
Operation progress 100 %
Disk information:
Size:
65537M
Minimum:
2338M
Minimum without resizing the last partition:
65523M
The information on the virtual disk is presented in the form of the following table:
Column Name
Description
Size
The virtual disk disk capacity, in megabytes, as it is seen from inside
the virtual machine.
Minimum
The virtual disk capacity, in megabytes, after resizing the disk using
the prl_disk_tool utility with the --resize_partition
option.
Minimum without
resizing the last
partition
The virtual disk capacity, in megabytes, after resizing the disk using
the prl_disk_tool utility without the --resize_partition
option.
Reducing the Disk Size
Once you know the minimum capacity of the virtual disk, you can start reducing it. For example, to
reduce the MyVM-0.hdd disk to 30 GB, you can execute the following command:
# prl_disk_tool resize --size 30G --hdd /vz/MyVM.pvm/MyVM-0.hdd --resize_partition
Operation progress 100 %
When reducing the disk capacity, keep in mind the following:
•
You cannot reduce the capacity of a virtual disk if the virtual machine using this disk is running.
•
The virtual machine using the virtual disk you want to configure must not have any snapshots.
Otherwise, you will be informed of this fact:
# prl_disk_tool resize --size 68000 --hdd /vz/MyVM.pvm/MyVM-0.hdd/
This disk has one or more snapshots and cannot be resized.
You need to delete snapshots using the pctl tool before resizing the disk.
90
Managing Virtual Machines and Containers
In this case, you should delete all existing snapshots and run the command again. To learn how
to delete virtual machine's snapshots, refer to Deleting a Snapshot (p. 85).
•
The capacity of an expanding virtual disk shown from inside the virtual machine and the size the
virtual disk occupies on the server's physical disk may differ.
Compacting Disks
In Parallels Server Bare Metal, you can decrease the space your virtual machines occupy on the
Parallels server's disk drive by compacting their virtual disks. Compacting virtual disks allows you to
save your server's disk space and host more virtual machines and Containers on the server.
Note: Plain disk cannot be compacted.
To compact a virtual disk, you can use the prl_disk_tool compact command. For example,
to compact the MyVM-0.hdd disk, you can run this command:
# prl_disk_tool compact --hdd /vz/MyVM.pvm/MyVM-0.hdd/
Operation progress 100 %
To check the space that was freed by compacting the virtual disk, you can use standard Linux
utilities (for example, the df utility).
91
Managing Virtual Machines and Containers
Managing Disk Interfaces
By default, any virtual machine is created with a SATA (Serial Advanced Technology Attachment)
virtual hard disk. If necessary, you can change the interface type of a disk from SATA to SCSI
(Small Computer System Interface) or IDE (Integrated Drive Electronics). For example, to change
the interface type of the default disk (hdd0) in the MyVM virtual machine from SATA to SCSI, you
can run the following command:
# pctl set MyVM --device-set hdd0 --iface scsi
The VM has been successfully configured
To check that the interface type has been successfully changed, use this command:
# pctl list -i MyVM | grep hdd0
Boot order: hdd0 cdrom0 fdd0 net0
hdd0 (+) scsi:0 image='/var/parallels/VM_SCSI.pvm/harddisk.hdd' 65536Mb
The command output shows that now the interface type of the hdd0 disk is SCSI.
You can create additional disks for the MyVM virtual machine. For example, to add a new disk of the
SCSI type to the virtual machine, execute the following command:
# pctl set MyVM --device-add hdd --iface scsi
Creating hdd1 (+) scsi:1 image='/var/parallels/MyVM.pvm/harddisk1.hdd' 65536Mb
Create the expanding image file, 65536Mb...
The VM has been successfully configured.
You can also create an IDE disk. To do this, specify --iface ide instead of --iface scsi in
the command above. If you omit the --iface option, a SATA disk is created by default.
The maximum number of devices (both virtual hard disks and CD/DVD-ROM drives) you can add to
a virtual machine is given below:
•
4 IDE devices
•
6 SATA devices
•
15 SCSI devices
At any time, you can remove the hdd1 disk from the MyVM virtual machine:
# pctl set MyVM --device-del hdd1
Remove the hdd1 device.
The VM has been successfully configured.
Notes:
1. SATA virtual disks can be added to or removed from both running and stopped virtual machines while
operations with IDE and SCSI disks can be performed on stopped virtual machines only.
2. You need to initialize a newly added disk before you can start using it. To initialize the disk, use
standard means provided by your guest operating system.
3. For more information on the pctl utility and its options, see the Parallels Server Bare Metal 5.0
Reference Guide.
92
Managing Virtual Machines and Containers
Managing Virtual Machine Devices
Parallels Server Bare Metal allows you to manage the following virtual machine devices:
•
hard disk drives
•
CD/DVD-ROM drives
•
floppy disk drives
•
network adapters
•
serial and parallels ports
•
sound cards
•
USB controllers
The main operations you can perform on these devices are:
•
adding a new device to the virtual machine
•
configuring the device properties
•
removing a device from the virtual machine
93
Managing Virtual Machines and Containers
Adding a New Device
This section provides information on adding new devices to your virtual machines. You can add
new virtual devices to your virtual machine using the pctl set command. The options
responsible for adding particular devices are listed in the following table:
Option Name
Description
hdd
Adds a new hard disk drive to the virtual machine. You can either
connect an existing image to the virtual machine or create a new
one.
Note: SATA disks can be added to running and stopped
virtual machines while IDE and SCSI disks—to stopped
virtual machines only.
cdrom
Adds a new CD/DVD-ROM drive to the virtual machine.
net
Adds a new network adapter to the virtual machine.
fdd
Adds a new floppy disk drive to the virtual machine.
serial
Adds a new serial port to the virtual machine.
parallel
Adds a new parallel port to the virtual machine.
sound
Adds a new sound device to the virtual machine.
usb
Adds a new USB controller to the virtual machine.
For example, you can execute the following command to add a new virtual disk to the MyVM virtual
machine:
# pctl set MyVM --device-add hdd
Creating hdd1 (+) sata:0 image='/var/parallels/MyVM.pvm/harddisk1.hdd
Create the expanding disk, 65536...
The VM has been successfully configured.
This command creates a new virtual disk with the following default parameters:
•
name: hdd1
•
disk type: SATA
•
image file name and location: /var/parallels/MyVM.pvm/harddisk1.hdd
•
disk format: expanding
•
disk capacity: 65536 MB
You can redefine some of these parameters by specifying specific options during the command
execution. For example, to create an IDE virtual disk that will have the capacity of 84 GB, you can
run this command:
# pctl set MyVM --device-add hdd --size 84000 --iface ide
Creating hdd1 (+) ide:1 image='/var/parallels/MyVM.pvm/harddisk1.hdd
Create the expanding disk, 84000Mb...
The VM has been successfully configured.
The virtual disk has been added to your virtual machine. However, before starting to use it, you
must initialize the disk. Refer to the next subsection for information on how you can do it.
94
Managing Virtual Machines and Containers
When managing devices, keep in mind the following:
•
Detailed information on all options that can be passed to pctl set when creating a new
virtual machine device is provided in the Parallels Server Bare Metal 5.0 Reference Guide.
•
You can connect up to 4 IDE devices, 6 SATA devices, and 15 SCSI devices (virtual disks or
CD/DVD-ROM drives) to a virtual machine.
•
If you want to use an existing image file as a virtual CD/DVD-ROM drive, keep in mind that
Parallels Server Bare Metal supports .iso, .cue, .ccd and .dmg (non-compressed and nonencrypted) image files.
•
A virtual machine can have only one floppy disk drive.
•
A virtual machine can have up to 16 virtual network adapters.
•
A virtual machine can have up to four serial ports.
•
A virtual machine can have up to three parallel ports.
•
Any virtual machine can have only one sound device.
•
A virtual machine can have only one USB controller.
95
Managing Virtual Machines and Containers
Initializing the Newly Added Disk
After you added a new blank virtual hard disk to the virtual machine configuration, it will be invisible
to the operating system installed inside the virtual machine until the moment you initialize it.
Initializing the New Virtual Hard Disk in Windows
To initialize a new virtual hard disk in a Windows guest OS, you will need the Disk Management
utility available. For example, in Windows 7 and Windows XP you can access this utility by doing
the following:
•
In Windows 7, click Start > Control Panel > System and Security > Administrative Tools >
Computer Management Storage > Disk Management.
•
In Windows XP, click Start > Control Panel > Administrative Tools > Computer
Management > Storage > Disk Management.
When you open the Disk Management utility, it automatically detects that a new hard disk was
added to the configuration and launches the Initialize and Convert Disk wizard:
1
In the introduction window, click Next.
2
In the Select Disks to Initialize window, select the newly added disk and click Next.
3
In the Select Disks to Convert window, select the newly added disk and click Finish.
The added disk will appear as a new disk in the Disk Management utility window, but its memory
space will be unallocated. To allocate the disk memory, right-click this disk name in the Disk
Management utility window and select New Simple Volume in Windows Vista or New Volume in
Windows XP. The New Simple Volume Wizard/New Volume Wizard window will appear. Follow
the steps of the wizard and create a new volume in the newly added disk.
After that your disk will become visible in Computer/My Computer and you will be able to use it as
a data disk inside your virtual machine.
Initializing the New Virtual Hard Disk in Linux
Initializing a new virtual hard disk in a Linux guest OS comprises two steps: (1) allocating the virtual
hard disk space and (2) mounting this disk in the guest OS.
To allocate the space, you need to create a new partition on this virtual hard disk using the fdisk
utility:
Note: To use the fdisk utility, you need the root privileges.
1
Launch a terminal window.
2
To list the IDE disk devices present in your virtual machine configuration, enter:
fdisk /dev/hd*
96
Managing Virtual Machines and Containers
Note: If you added a SCSI disk to the virtual machine configuration, use the fdisk /dev/sd*
command instead.
3
By default, the second virtual hard disk appears as /dev/hdc in your Linux virtual machine. To
work with this device, enter:
fdisk /dev/hdc
Note: If this is a SCSI disk, use the fdisk /dev/sdc command instead.
4
To get detailed information about the disk, enter:
p
5
To create a new partition, enter:
n
6
To create the primary partition, enter:
p
7
Specify the partition number. By default, it is 1.
8
Specify the first cylinder. If you want to create a single partition on this hard disk, use the default
value.
9
Specify the last cylinder. If you want to create a single partition on this hard disk, use the default
value.
10 To create a partition with the specified settings, enter:
w
When you allocated the space on the newly added virtual hard disk, you should format it by
entering the following command in the terminal:
mkfs -t <FileSystem> /dev/hdc1
Note: <FileSystem> stands for the file system you want to use on this disk. It is recommended to
use ext3 or ext2.
When the added virtual hard disk is formatted, you can mount it in the guest OS.
1
To create a mount point for the new virtual hard disk, enter:
mkdir /mnt/hdc1
Note: You can specify a different mount point.
2
To mount the new virtual hard disk to the specified mount point, enter:
mount /dev/hdc1 /mnt/hdc1
When you mounted the virtual hard disk, you can use its space in your virtual machine.
97
Managing Virtual Machines and Containers
Configuring Virtual Devices
In Parallels Server Bare Metal, you can use the --device-set option of the pctl set
command to configure the parameters of an existing virtual device. As a rule, the process of
configuring the device properties includes two steps:
1
Finding out the name of the device you want to configure.
2
Running the pctl set command to configure the necessary device properties.
Finding Out Device Names
To configure a virtual device, you need to specify its name when running the pctl set command.
If you do not know the device name, you can use the pctl list command to learn it. For
example, to obtain the list of virtual devices in the MyVM virtual machine, run this command:
# pctl list --info MyVM
...
Hardware:
cpu 2 VT-x accl=high mode=32
memory 256Mb
video 46Mb
fdd0 (+) real='/dev/fd0' state=disconnected
hdd0 (+) sata:0 image='/var/parallels/MyVM.pvm/harddisk.hdd' 27000Mb
hdd1 (+) scsi:0 image='/var/parallels/MyVM.pvm/harddisk1.hdd' 32768Mb
cdrom0 (+) ide:1 real='Default CD/DVD-ROM'
parallel0 (+) real='/dev/lp0'
usb (+)
net0 (+) type=bridged iface='eth1' mac=001C4201CED0
...
All virtual devices currently available to the virtual machine are listed under Hardware. In our case
the MyVM virtual machine has the following devices: 2 CPUs, main memory, video memory, a floppy
disk drive, 2 hard disk drives, a CD/DVD-ROM drive, a parallel port, a USB controller, and a
network card.
Configuring Virtual Devices
Once you know the virtual device name, you can configure its properties. For example, you can
execute the following command to configure the current type of the virtual disk hdd1 in the MyVM
virtual machine from SATA to SCSI:
# pctl set MyVM --device-set hdd1 --iface scsi
The VM has been successfully configured.
To check that the virtual disk type has been successfully changed, use the pctl list --info
command:
# pctl list --info MyVM
...
hdd0 (+) scsi:1 image='/var/parallels/MyVM.pvm/harddisk.hdd' 85000Mb
...
98
Managing Virtual Machines and Containers
Connecting and Disconnecting Virtual Devices
In Parallels Server Bare Metal, you can connect or disconnect certain devices when a virtual
machine is running. These devices include:
•
SATA hard drives
•
CD/DVD-ROM drives
•
floppy disk drives
•
network adapters
•
printer ports
•
serial ports
•
sound devices
•
USB devices
•
shared folders
Usually, all virtual devices are automatically connected to a virtual machine when you create them.
To disconnect a device from the virtual machine, you can use the pctl set command. For
example, the following command disconnects the CD/DVD-ROM drive cdrom0 from the MyVM
virtual machine:
# pctl set MyVM --device-disconnect cdrom0
Disconnect device: cdrom0
The VM has been successfully configured.
To connect the CD/DVD-ROM drive back, you can run the following command:
# pctl set MyVM --device-connect cdrom0
Connect device: cdrom0
The VM has been successfully configured.
99
Managing Virtual Machines and Containers
Deleting a Device
You can delete a virtual device that you do not need any more in your virtual machine using the -device-del option of the pctl set command. The options responsible for removing particular
devices are listed in the following table:
Option Name
Description
hdd
Deletes the specified hard disk drive from the virtual machine.
Note: SATA disks can be removed from running and
stopped virtual machines while IDE and SCSI disks—from
stopped virtual machines only.
cdrom
Deletes the specified CD/DVD-ROM drive from the virtual machine.
net
Deletes the specified network adapter from the virtual machine.
fdd
Deletes the floppy disk drive from the virtual machine.
serial
Deletes the specified serial port from the virtual machine.
parallel
Deletes the specified parallel port from the virtual machine.
sound
Deletes the sound device from the virtual machine.
usb
Deletes the USB controller from the virtual machine.
As a rule deleting a virtual device involves performing two operations:
1
Finding out the name of the device to be deleted.
2
Deleting the device from the virtual machine.
Finding Out the Device Name
To remove a virtual device, you need to specify its name when running the pctl set command. If
you do not know the device name, you can use the pctl list command to learn it. For example,
to obtain the list of virtual devices in the MyVM virtual machine, run this command:
# pctl list --info MyVM
...
Hardware:
cpu 2 VT-x accl=high mode=32
memory 256Mb
video 46Mb
fdd0 (+) real='/dev/fd0' state=disconnected
hdd0 (+) ide:0 image='/var/parallels/MyVM.pvm/harddisk.hdd' 27Mb
hdd1 (+) scsi:0 image='/var/parallels/MyVM.pvm/harddisk1.hdd' 32768Mb
cdrom0 (+) ide:1 real='Default CD/DVD-ROM'
parallel0 (+) real='/dev/lp0'
usb (+)
net0 (+) type=bridged iface='eth1' mac=001C4201CED0
...
100
Managing Virtual Machines and Containers
All virtual devices currently available to the virtual machine are listed under Hardware. In our case
the MyVM virtual machine has the following devices: 2 CPUs, main memory, video memory, a floppy
disk drive, 2 hard disk drives, a CD/DVD-ROM drive, a parallel port, a USB controller, and a
network card.
Deleting a Virtual Device
Once you know the virtual device name, you can remove it from your virtual machine. For example,
you can execute the following command to remove the virtual disk hdd1 from the MyVM virtual
machine:
# pctl set MyVM --device-del hdd1
Remove the hdd1 device.
The VM has been successfully configured.
When deleting virtual machine devices, keep in mind the following:
•
If you do not want to permanently delete a virtual device, you can temporarily disconnect if from
the virtual machine using the --disable option.
•
Detailed information on all options that can be used with pctl set when deleting a device is
given in the Parallels Command Line Reference Guide.
Making Screenshots
In Parallels Server Bare Metal, you can use the pctl capture command to capture an image (or
screenshot) of your virtual machine screen. You can take screenshots of running virtual machines
only. The session below demonstrates how to take a screenshot of the My_VM virtual machine
screen and save it to the /usr/screenshots/image1.png file:
1
Make sure that the virtual machine is running:
# pctl list
STATUS
IP_ADDR
NAME
running 10.10.10.101 101
running
10.10.10.201 My_VM
2
Take the virtual machine screenshot:
# pctl capture My_VM --file /usr/screenshots/image1.png
Capture the VM screen...
The VM screen has been saved to /usr/screenshots/image1.png
3
Check that the image1.png file has been successfully created:
# ls /usr/screenshots/
image1.png
101
Managing Virtual Machines and Containers
Assigning USB Devices to Virtual Machines
In Parallels Server Bare Metal, you can assign a USB device to a virtual machine so that the device
is automatically connected to the virtual machine when you connect the USB device to the Parallels
server or start the virtual machine. To assign a USB device to a virtual machine, you need to specify
two parameters:
•
ID of the USB device. To get this information, use the prlsrvctl info command, for
example:
# prlsrvctl info
...
Hardware info:
hdd
hdd-part NTFS
hdd-part Linux
hdd-part Linux
hdd-part Linux swap
cdrom Optiarc DVD RW AD-7260S
net eth0
usb Broadcom - USB Device 3503
usb Broadcom - USB Device 3502
usb LITEON Technology - USB Multimedia Keyboard
serial /dev/ttyS0
serial /dev/ttyS1
'/dev/sda'
'/dev/sda2'
'/dev/sda3'
'/dev/sda5'
'/dev/sda6'
'/dev/scd0'
'eth0'
'2-1.4.3|0a5c|3503|full|KM|Empty'
'2-1.4.2|0a5c|3502|full|KM|Empty'
'1-1.6|046d|c312|low|KM|Empty'
'/dev/ttyS0'
'/dev/ttyS1'
All USB devices available on the Parallels server are listed in the Hardware info section and
start with usb.
•
ID of the virtual machine. To get this information, use the pctl list --info command, for
example:
# pctl list --info
ID: {d8d516c9-dba3-dc4b-9941-d6fad3767035}
Name: Windows 7
...
The first line in the command output indicates the virtual machine ID; in our case, it is
{d8d516c9-dba3-dc4b-9941-d6fad3767035}.
Once you know the USB device and virtual machine IDs, you can use the prlsrvctl usb set
command to assign the USB device to the virtual machine. For example:
# prlsrvctl usb set '1-1.6|046d|c312|low|KM|Empty' {d8d516c9-dba3-dc4b-9941d6fad3767035}
The server has been successfully configured.
This command assigns the USB device LITEON Technology - USB Multimedia
Keyboard with ID '1-1.6|046d|c312|low|KM|Empty' to the virtual machine with ID
{d8d516c9-dba3-dc4b-9941-d6fad3767035}. When running the command, remember to
specify the single quotes and curly brackets with the USB device and virtual machine IDs,
respectively.
To check that the USB device has been successfully assigned to the virtual machine, use the
prlsrvctl usb list command:
# prlsrvctl usb list
102
Managing Virtual Machines and Containers
Broadcom - USB Device 3503
'2-1.4.3|0a5c|3503|full|KM|Empty'
Broadcom - USB Device 3502
'2-1.4.2|0a5c|3502|full|KM|Empty'
LITEON Technology - USB Multimedia Keyboard '1-1.6|046d|c312|low|KM|Empty'
{d8d516c9-dba3-dc4b-9941-d6fad3767035}
The command output shows that the USB device with ID '11.6|046d|c312|low|KM|Empty' is now associated with the virtual machine with ID
{d8d516c9-dba3-dc4b-9941-d6fad3767035}. This means that the device is automatically
connected to the virtual machine every time you start this virtual machine and connect the device to
the Parallels server.
To remove the assignment of the USB device with ID '1-1.6|046d|c312|low|KM|Empty',
use the prlsrvctl usb del command:
# prlsrvctl usb del '1-1.6|046d|c312|low|KM|Empty'
The server has been successfully configured.
When assigning USB devices to virtual machines, keep in mind the following:
•
You cannot migrate a running virtual machine having one or more USB devices assigned.
•
After migrating a stopped virtual machine, all its assignments are lost.
•
All USB assignments are preserved if you restoring a virtual machine to its original location and
are lost otherwise.
•
The USB device assignment and a virtual machine is created for the user currently logged in to
the system.
103
Managing Virtual Machines and Containers
Mounting Virtual Machines
In Parallels Server Bare Metal, you can mount virtual machines using the pctl mount command.
Once you mount a virtual machine, its contents appears in the /vz/mnt/VM_ID directory on the
Parallels server. You can then explore this directory and manage the necessary files and
subdirectories. Let us assume that you want to mount the MyVM virtual machine that
•
runs Windows XP
•
has the ID of {36ecc3c5-ec50-6044-8c59-241ea5d7183d}
•
contains two disks (C: and D:)
To mount the MyVM virtual machine, run the following command:
# pctl mount MyVM
Mounting the VM...
The VM has been successfully mounted.
Once the virtual machine is successfully mounted, the directory /vz/mnt/{36ecc3c5-ec506044-8c59-241ea5d7183d} appears on the Parallels server. As the MyVM virtual machine has
two disks, this directory contains two subdirectories: volume_1 corresponding to the C: disk of
the virtual machine and volume_2 corresponding to the D: disk. By default, the disks are
mounted in the read-write mode; so you can go to any of the directories and view or change their
contents as necessary. You can also mount the disks in the read-only mode by using the -o
option:
# pctl mount MyVM -o ro
Mounting the VM...
The VM has been successfully mounted.
You can use the --info option of the pctl mount command to view the information about the
mounted disks:
# pctl mount MyVM --info
Volume 1:
/vz/vmprivate/MyVM.pvm/config.pvs
filesystem
NTFS
size
63 G
free space
61 G
ID: /vz/mnt/{36ecc3c5-ec50-6044-8c59-241ea5d7183d}/volume_1
Volume 2: ...
This command outputs the following information for each mounted disk:
•
full path to the virtual machine configuration file
•
file system on the disk
•
free space available on the disk
•
full path to the mounted disk
Note: To start a mounted virtual machine, you need to unmount it first.
104
Managing Virtual Machines and Containers
Configuring IP Address Ranges for Host-Only Networks
All virtual machines connected to networks of the host-only type receive their IP addresses from the
Parallels DHCP server. This DHCP server is set up during the Parallels Server Bare Metal installation
and includes by default IP addresses from 10.37.130.1 to 10.37.130.254. You can redefine the
default IP address range for host-only networks and make virtual machines get their IP addresses
from different IP address ranges. For example, you can run the following command to set the start
and end IP addresses for the Host-Only network (this network is automatically created during the
Parallels Server Bare Metal installation) to 10.10.11.1 and 10.10.11.254, respectively:
# prlsrvctl net set Host-Only --ip-scope-start 10.10.11.1 --ip-scope-end 10.10.11.254
You can also specify a custom IP address range directly when creating a new network of the hostonly type. Assuming that you want to create a network with the Host-Only2 name and define for
this network the IP addresses range from 10.10.10.1 to 10.10.10.254, you can execute the
following command:
# prlsrvctl net add Host-Only2 -t host-only --ip-scope-start 10.10.10.1 --ip-scope-end
10.10.10.254
When working with IP address ranges, pay attention to the following:
•
The start and end IP addresses of an IP address range must belong to the same subnetwork.
•
IP address ranges can be defined for each network of the host-only type separately. For
example, you can set the IP address range from 10.10.11.1 to 10.10.11.254 for the HostOnly network and from 10.10.10.1 to 10.10.10.254 for the Host-Only2 network.
105
Managing Virtual Machines and Containers
Converting Third-Party Virtual Machines and Disks
In Parallels Server Bare Metal, you can convert third-party virtual machines and their disks to
Parallels virtual machines and disks. Currently, you can convert the following third-party virtual
machines and disks:
•
Microsoft Hyper-V
•
Microsoft Virtual PC
•
Virtual Box
•
VMware
Converting Virtual Machines
Let us assume that you want to convert a VMware virtual machine that runs the CentOS 5
operating system and has the name centos5. As the pctl convert command can work only
with virtual machines and disks that are available locally, you first need to copy the virtual machine
to the Parallels server. Once the virtual machine is on your local server, you can start the
conversion. Assuming that you have copied the virtual machine to the /var/parallels directory
on the Parallels server and the full path to its configuration file is
/var/parallels/centos5/config.xml, you can run the following command to perform the
conversion:
# pctl convert /var/parallels/centos5/config.xml
Once the conversion is complete, you can start the virtual machine and manage it in the same way
you would manage a native Parallels virtual machine.
Converting Disks
You can also convert third-party virtual disks to Parallels virtual machines and disks using the
prl_convert utility. Once you run the utility, it checks the disk and, depending on its type, does
one of the following:
•
If the disk is a system disk—that is, has an operating system installed, prl_convert converts
it to a Parallels virtual machine. If the utility cannot create a virtual machine for the disk (for
example, it fails to detect the operating system on the disk), the disk is converted to a Parallels
virtual disk. You can also specify the --allow-no-os option to force the conversion, but in
this case you may have problems with starting and using the resulting virtual machine.
•
If the disk is a data disk, prl_convert converts it to a Parallels virtual disk.
When converting third-party virtual disks, you need to specify the full path to the original disk file.
For example, to convert the system disk of the centos5 virtual machine (that is, the disk where the
CentOS 5 operating system is installed) that has the full path of
/var/parallels/centos5/centos5.vhd, you can use this command:
# prl_convert /var/parallels/centos5/centos5.vhd
106
Managing Virtual Machines and Containers
This command creates a ready-to-use Parallels virtual machine with the name centos5. You can
start this virtual machine and manage it in the same way you would manage a native Parallels virtual
machine. At the same time, if you convert a third-party virtual data disk, you will need first to add
the resulting disk to an existing Parallels virtual machine using the pctl set command.
Notes:
1. When adding a converted virtual disk to an existing Parallels virtual machine or creating a new virtual
machine on its basis, make sure that the interface type of the disk is the same as it was in the source
virtual machine.
For example, if the original disk had the SCSI interface type, ensure that the interface type of the
converted disk is also set to SCSI. If you do not configure the disk interface type, it will be set to SATA
(this is the default interface type in Parallels virtual machines), which may cause your virtual machine to
malfunction.
2. In the current version of Parallels Server Bare Metal, Hyper-V virtual machines can be converted using
the prl_convert utility only. That means that you first need to convert all Hyper-V virtual disks and then
add them to an existing virtual machine.
107
Managing Virtual Machines and Containers
Enabling VNC Access to Virtual Machines
By default, you cannot connect to a virtual machine using a VNC client. To enable VNC access to a
virtual machine, you need to configure the following settings:
1
Enable the VNC support in the virtual machine.
2
Specify the TCP port number on the physical server that will be used to listen to VNC
connections for the virtual machine.
Note: A unique port number must be specified for each virtual machine where you plan to connect via
VNC.
3
Set a password to secure your VNC connection. Anyone trying to connect to the virtual
machine will need to specify this password before they will be logged in to the virtual machine.
Let us assume the following:
•
You want to enable VNC access to the MyVM virtual machine.
•
The server will listen on port 5901 for VNC connections to the virtual machine.
•
The XXXXXXXXX password will be used to access the virtual machine.
To enable VNC access to the MyVM virtual machine and set the parameters above, you can run the
following command:
# pctl set MyVM --vnc-mode manual --vnc-port 5901 --vnc-passwd XXXXXXXXX
Configure VNC: mode=manual port=5901
Success. The operation was successfully completed.
The changes will come into effect on the next virtual machine start.
108
CHAPTER 4
Managing Resources
The main goal of resource control in Parallels Server Bare Metal is to provide Service Level
Management or Quality of Service for virtual machines and Containers. Correctly configured
resource control settings prevent serious impacts resulting from the resource over-usage
(accidental or malicious) of any virtual machine and Container on the other virtual machines and
Containers. Using resource control parameters for resources management also allows you to
enforce fairness of resource usage among virtual machines and Containers and better service
quality for preferred virtual machines and Containers, if necessary.
In This Chapter
What are Resource Control Parameters? ................................................................ 109
Managing CPU Resources...................................................................................... 110
Managing Disk Quotas ........................................................................................... 115
Managing Network Accounting and Bandwidth ....................................................... 125
Managing Disk I/O Parameters ............................................................................... 134
Managing Memory Parameters for Containers ......................................................... 139
Managing Memory Resources for Virtual Machines.................................................. 144
Managing Container Resources Configuration ......................................................... 149
What are Resource Control Parameters?
The system administrator can control the resources available to a virtual machine and Container
through a set of resource management parameters. All these parameters can be set and
configured as follows:
•
For Containers, by manually editing the global (/etc/vz/vz.conf) and Container
(/etc/vz/conf/CT_ID) configuration files or using the Parallels Server Bare Metal commandline utilities.
•
For virtual machines, by manually editing virtual machine configuration files
(/etc/vz/VM_Name.pvm/config.pvs) or using the Parallels Server Bare Metal commandline utilities.
The following sections describe in detail how to configure resource parameters for both virtual
machines and Containers.
Managing Resources
Managing CPU Resources
This section explains the CPU resource parameters that you can configure and monitor for each
virtual machine and Container. The table below provides the name and the description for the CPU
parameters.
Parameter
Description
cpuunits
This is a positive integer number that determines the minimal
guaranteed share of the CPU time the virtual machine or Container will
receive.
cpulimit
This is a positive number indicating the CPU time, in percent, the
virtual machine or Container is not allowed to exceed.
cpus
The number of CPUs to be used to handle the processes running
inside the virtual machine or Container.
110
Managing Resources
Managing CPU Shares and Limits
The Parallels Server Bare Metal CPU resource control utilities allow you to guarantee any virtual
machine or Container the amount of CPU time this virtual machine or Container receives. The
virtual machine or Container can consume more than the guaranteed value if no other virtual
machines and Containers are competing for the CPU and the cpulimit parameter is not defined.
Note: The CPU time shares and limits are calculated on the basis of a one-second period. Thus, for
example, if a virtual machine or Container is not allowed to receive more than 50% of the CPU time, it will
be able to receive no more than half a second each second.
To get a view of the optimal share to assign to a virtual machine or Container, check the current
server CPU utilization:
# vzcpucheck
Current CPU utilization: 11142
Power of the node: 125504
The output of this command displays the total number of the so-called CPU units consumed by all
running virtual machines and Containers and server processes. This number is calculated by
Parallels Server Bare Metal with the help of a special algorithm. The above example illustrates the
situation when the server is underused. In other words, the running virtual machines and Containers
can receive more CPU time than is guaranteed to them.
In the following example, the MyVM virtual machine is guaranteed to receive about 4% of the CPU
time even if the server is fully used—that is, if the current CPU utilization equals the power of the
server. Besides, this virtual machine will not receive more than 25% of the CPU time even if the
CPU is not fully loaded:
# pctl set MyVM --cpuunits 5000 --cpulimit 25
set cpuunits 5000
set cpulimit 25%
The VM has been successfully configured.
To set the CPU time share and limit for Containers, just specify a Container ID instead of the virtual
machine name and use the --save option to save the changes in the Container configuration file,
for example:
# pctl set 101 --cpuunits 5000 --cpulimit 25 --save
Saved parameters for Container 101
Setting CPU limits in megahertz
In the example above, you set the CPU limit for Container 101 and the MyVM virtual machine to
25% of the whole CPU power of the Parallels server. That means that if the server CPU power is
2000 megahertz (MHz), Container 101 and the MyVM virtual machine can get up to 500 MHz each.
111
Managing Resources
Now, imagine the situation when you migrate Container 101 to another server whose CPU power
equals 4000 MHz. On this server, Container 101 can get 25% of 4000 MHz—that is, 1000 MHz. To
deal with such situations, you can set CPU limits for virtual machines and Containers in MHz. Once
you set the CPU limit in MHz, it remains the same, irrespective of the server total CPU power. For
example, to make Container 101 and the MyVM virtual machine consume no more than 500 MHz
on any Parallels server, you can run the following commands:
# pctl set MyVM ---cpulimit 500m
set cpulimit 500 MHz
The VM has been successfully configured.
# pctl set 101 --cpulimit 500m --save
Saved parameters for Container 101
112
Managing Resources
Configuring the Number of CPUs
If your server has more than one physical processor installed, you can control the number of CPUs
which will be used to handle the processes running in virtual machines and Containers. By default,
a virtual machine and Container is allowed to consume the CPU time of all processors on the
server, i.e. any process in any virtual machine and Container can be executed on any processor on
the server. However, you can modify the number of physical CPUs which will be simultaneously
available to a virtual machine or Container using the --cpus option of the pctl set command.
For example, if your server has 4 physical processors installed, i.e. any virtual machine and
Container on the server can make use of these 4 processors, you can set the processes in
Container 101 and in the MyVM virtual machine to be run on 2 CPUs only by issuing the following
commands:
# pctl set 101 --cpus 2 --save
Saved parameters for Container 101
# pctl set MyVM --cpus 2
set cpus(4): 2
The VM has been successfully configured.
Notes:
1. The number of CPUs for a virtual machine or Container must not exceed the number of physical CPUs
installed on the server. In this case, the 'physical CPUs' notation designates the number of CPUs the
Parallels Server Bare Metal kernel is aware of (you can view this CPU number using the
/proc/cpuinfo command).
2. Before configuring the number of CPUs for a running virtual machine, make sure that the CPU hotplug
support is enabled for this virtual machine. For detailed information on the CPU hotplug feature, see
Enabling CPU Hotplug for Virtual Machines (p. 114).
To make sure that the number of CPUs has been successfully changed, you can log in to
Container 101 and the MyVM virtual machine and check the number of available CPUs. If the
Container and virtual machine are running a Linux operating system, you can also use the cat
/proc/cpuinfo command to view the available CPUs, for example:
# pctl exec 101 cat /proc/cpuinfo
processor
: 0
vendor_id
: GenuineIntel
cpu family
: 15
model
: 4
model name
: Intel(R) Xeon(TM) CPU 2.80GHz
stepping
: 1
cpu MHz
: 2793.581
cache size
: 1024 KB
...
processor
vendor_id
cpu family
model
model name
stepping
cpu MHz
cache size
:
:
:
:
:
:
:
:
1
GenuineIntel
15
4
Intel(R) Xeon(TM) CPU 2.80GHz
1
2793.581
1024 KB
113
Managing Resources
...
The output shows that Container 101 is currently bound to only two processors on the server
instead of 4 available for the other Containers on this server. It means that, from this point on, the
processes of Container 101 will be simultaneously executed on no more than 2 physical CPUs
while the other Containers on the server will continue consuming the CPU time of all 4 server
processors, if needed. Also notice that the physical CPUs proper of Container 101 might not
remain the same during the Container operation; they might change for load balancing reasons, the
only thing that cannot be changed is their maximal number.
Enabling CPU Hotplug for Virtual Machines
If a guest operating system supports the CPU hotplug functionality, you can enable this functionality
for the virtual machine. Once the CPU hotplug functionality is turned on, you can configure the
number of CPUs available to your virtual machines even if they are running. Currently, the following
systems come with the CPU hotplug support:
Linux (both x86 and x64 versions)
•
Linux operating systems based on the RHEL 5 kernel and higher (Red Hat Linux Enterprise 5,
CentOS 5, and so on)
Windows
•
x64 version of Windows Server 2008 (Standard Edition)
•
x64 version of Windows Server 2008 (Enterprise Edition)
•
x64 version of Windows Server 2008 (Datacenter Edition)
•
x64 version of Windows Server 2008 R2 (Datacenter Edition)
By default, the CPU hotplug support is disabled for all newly created virtual machines. To enable
this functionality, you can use the --cpu-hotplug option of the pctl set command. For
example, to enable the CPU hotplug support in the MyVM virtual machine that runs one of the
supported operating systems, stop the MyVM virtual machine and run this command:
# pctl set MyVM --cpu-hotplug on
set cpu hotplug: 1
The VM has been successfully configured.
Once the functionality is enabled, you can configure the number of CPUs in the MyVM virtual
machine even it is running. Assuming that your physical server has 4 CPUs installed and the
processes in the MyVM virtual machine are set to be executed on two CPUs, you can run the
following command to assign 3 CPUs to the virtual machine:
# pctl set MyVM --cpus 3
set cpus(4): 3
The VM has been successfully configured.
To disable the CPU hotplug support in the MyVM virtual machine, use this command:
# pctl set MyVM --cpu-hotplug off
set cpu hotplug: 0
The VM has been successfully configured.
The changes will come into effect on the next virtual machine start.
114
Managing Resources
Configuring CPU Affinity for Virtual Machines and Containers
If your physical server has several processors installed, you can bind a virtual machine or Container
to specific CPUs so that only these CPUs are used to handle the processes running in the virtual
machine or Container. The feature of binding certain processes to certain CPUs is known as CPU
affinity. Establishing CPU affinity between virtual machines and Containers and physical processors
can help you increase your system performance up to 20%.
By default, any newly created virtual machine and Container can consume the CPU time of all
processors installed on the physical server. To bind a virtual machine or Container to specific
CPUs, you can use the --cpumask option of the pctl set command. Assuming that your
physical server has 8 CPUs, you can make the processes in the MyVM virtual machine and
Container 101 run on CPUs 0, 1, 3, 4, 5, and 6 by running the following commands:
# pctl set MyVM --cpumask 0,1,3,4-6
The VM has been successfully configured.
# pctl set 101 --cpumask 0,1,3,4-6 --save
Saved parameters for Container 101
You can specify the CPU affinity mask—that is, the processors to bind to virtual machines and
Containers—as separate numbers (0,1,3) or as ranges (4-6). If you are setting the CPU affinity
mask for a running virtual machine or Container, the changes are applied on the fly.
To undo the changes made to the MyVM virtual machine and Container 101 and set their processes
to run on all available CPUs on the server, run these commands:
# pctl set MyVM --cpumask all
The VM has been successfully configured.
# pctl set 101 --cpumask all --save
Saved parameters for Container 101
Managing Disk Quotas
This section explains the basics of disk quotas, defines disk quota parameters, and describes how
to perform the following disk quota related operations:
•
turning on and off per-Container (first-level) disk quotas
•
setting up first-level disk quota parameters for a Container
•
turning on and off per-user and per-group (second-level) disk quotas inside a Container
•
setting up second-level quotas for a user or for a group
•
checking disk quota statistics
•
cleaning up Containers
115
Managing Resources
What are Disk Quotas?
Disk quotas enable system administrators to control the size of Linux file systems by limiting the
amount of disk space and the number of inodes a Container can use. These quotas are known as
per-Container quotas or first-level quotas in Parallels Server Bare Metal. In addition, the Parallels
Server Bare Metal software enables the Container administrator to limit disk space and the number
of inodes that individual users and groups in that Container can use. These quotas are called peruser and per-group quotas or second-level quotas.
By default, first-level quotas on your server are enabled (which is defined in the
/etc/vz/vz.conf configuration file), whereas second-level quotas must be turned on for each
Container separately (in the corresponding Container configuration files). It is impossible to turn on
second-level disk quotas for a Container if first-level disk quotas are off for that Container.
Parallels Server Bare Metal keeps quota usage statistics and limits in
/var/vzquota/quota.<CT_ID> - a special quota file. The quota file has a special flag
indicating whether the file is “dirty”. The file becomes dirty when its contents become inconsistent
with the real Container usage. This means that when the disk space or inodes usage changes
during the Container operation, these statistics are not automatically synchronized with the quota
file, the file just gets the “dirty” flag. They are synchronized only when the Container is stopped or
when the server is shut down. After synchronization, the “dirty” flag is removed. If the server has
been incorrectly brought down (for example, the power switch was hit), the file remains “dirty”, and
the quota is re-initialized on the next Container startup. This operation may noticeably increase the
server startup time. Thus, it is highly recommended to shut down the server properly.
Disk Quota Parameters
The table below summarizes the disk quota parameters that you can control. The File column
indicates whether the parameter is defined in the global configuration file (G), in the Container
configuration files (V), or it is defined in the global configuration file but can be overridden in a
separate Container configuration file (GV).
Parameter
Description
DISK_QUOTA
Indicates whether first-level quotas are on or off for all Containers or for an GV
individual Container.
DISKSPACE
Total size of disk space the Container may consume, in 1-Kb blocks.
V
DISKINODES
Total number of disk inodes (files, directories, and symbolic links) the
Container can allocate.
V
QUOTATIME
Grace period for the disk quota overusage, in seconds. The Container is
allowed to temporarily exceed its quota soft limits for no more than the
QUOTATIME period.
V
QUOTAUGIDLIMIT
Maximum aggregate number of user IDs and group IDs for which disk
V
quota inside the given Container will be accounted. If set to 0, the UID and
GID quota are disabled.
116
File
Managing Resources
Turning On and Off Per-Container Disk Quotas
The parameter that defines whether to use first-level disk quotas is DISK_QUOTA in the global
configuration file (/etc/vz/vz.conf). By setting it to “no”, you will disable disk quotas
completely.
This parameter can be specified in the Container configuration file
(/etc/vz/conf/<CT_ID>.conf) as well. In this case, its value will take precedence of the one
specified in the global configuration file. If you intend to have a mixture of Containers with quotas
turned on and off, it is recommended to set the DISK_QUOTA value to yes in the global
configuration file and to no in the configuration file of that Container which does not need quotas.
The session below illustrates a scenario when first-level quotas are on by default and are turned off
for Container 101:
[checking that quota is on]
# grep DISK_QUOTA /etc/vz/vz.conf
DISK_QUOTA=yes
[checking available space on /vz partition]
# df /vz
Filesystem
1k-blocks
Used Available Use% Mounted on
/dev/sda2
8957295
1421982
7023242 17% /vz
[editing Container configuration file to add DISK_QUOTA=no]
# vi /etc/vz/conf/101.conf
[checking that quota is off for Container 101]
# grep DISK_QUOTA /etc/vz/conf/101.conf
DISK_QUOTA=no
# pctl start 101
Starting Container ...
Container is mounted
Adding IP address(es): 10.0.16.101
Hostname for Container set: ve101
Container start in progress...
# pctl exec 101 df
Filesystem
1k-blocks
Used Available Use% Mounted on
vzfs
8282373
747060
7023242 10% /
As the above example shows, the only disk space limit a Container with the quotas turned off has
is the available space and inodes on the partition where the Container private area resides.
117
Managing Resources
Setting Up Per-Container Disk Quota Parameters
Three parameters determine how much disk space and inodes a Container can use. These
parameters are specified in the Container configuration file:
DISKSPACE
The total size of disk space that can be consumed by the Container, in 1-Kb blocks.
When the space used by the Container hits the soft limit, the Container can allocate
additional disk space up to the hard limit during the grace period specified by the
QUOTATIME parameter.
DISKINODES
The total number of disk inodes (files, directories, and symbolic links) the Container
can allocate. When the number of inodes used by the Container hits the soft limit,
the Container can create additional file entries up to the hard limit during the grace
period specified by the QUOTATIME parameter.
QUOTATIME
The grace period of the disk quota, in seconds. The Container is allowed to
temporarily exceed the soft limit values for the disk space and disk inodes quotas for
no more than the period specified by this parameter.
The first two parameters have both soft and hard limits (or, simply, barriers and limits). The hard
limit is the limit that cannot be exceeded under any circumstances. The soft limit can be exceeded
up to the hard limit, but as soon as the grace period expires, the additional disk space or inodes
allocations will fail. Barriers and limits are separated by colons (“:”) in Container configuration files
and in the command line.
The following session sets the disk space available to Container 101 to approximately 1 GB and
allows the Container to allocate up to 90,000 inodes. The grace period for the quotas is set to 10
minutes:
# pctl set 101 --diskspace 1000000:1100000 --save
Saved parameters for Container 101
# pctl set 101 --diskinodes 90000:91000 --save
Saved parameters for Container 101
# pctl set 101 --quotatime 600 --save
Saved parameters for Container 101
# pctl exec 101 df
Filesystem
1k-blocks
Used Available Use% Mounted on
vzfs
1000000
747066
252934 75% /
# pctl exec 101 stat -f /
File: "/"
ID: 0
0
Namelen: 255
Type: UNKNOWN (0x565a4653)
Blocks: Total: 1000000
Free: 252934
Available: 252934
Size: 1024
Inodes: Total: 90000
Free: 9594
It is possible to change the first-level disk quota parameters for a running Container. The changes
will take effect immediately. If you do not want your changes to persist till the next Container
startup, do not use the –-save switch.
118
Managing Resources
Turning On and Off Second-Level Quotas for a Container
The parameter that controls the second-level disk quotas is QUOTAUGIDLIMIT in the Container
configuration file. By default, the value of this parameter is zero and this corresponds to disabled
per-user and per-group quotas.
If you assign a non-zero value to the QUOTAUGIDLIMIT parameter, this action brings about the
two following results:
1
Second-level (per-user and per-group) disk quotas are enabled for the given Container.
2
The value that you assign to this parameter will be the limit for the number of file owners and
groups of this Container, including Linux system users. Notice that you will theoretically be able
to create extra users of this Container, but if the number of file owners inside the Container has
already reached the limit, these users will not be able to own files.
Enabling per-user and per-group quotas for a Container requires restarting the Container. The
value for it should be carefully chosen; the bigger value you set, the bigger kernel memory overhead
this Container creates. This value must be greater than or equal to the number of entries in the
Container /etc/passwd and /etc/group files. Taking into account that a newly created Red
Hat Linux-based Container has about 80 entries in total, the typical value would be 100. However,
for Containers with a large number of users, this value should be increased.
When managing the QUOTAUGIDLIMIT parameter, keep in mind the following:
•
If you delete a registered user but some files with their ID continue residing inside your
Container, the current number of ugids (user and group identities) inside the Container will not
decrease.
•
If you copy an archive containing files with user and group IDs not registered inside your
Container, the number of ugids inside the Container will increase by the number of these new
IDs.
The session below turns on second-level quotas for Container 101:
# pctl set 101 --quotaugidlimit 100 --save
Unable to apply new quota values: ugid quota not initialized
Saved parameters for Container 101
# pctl restart 101
Stopping Container ...
Container was stopped
Container is unmounted
Starting Container ...
Container is mounted
Adding IP address(es): 192.168.1.101
Hostname for Container set: ct101
Container start in progress...
119
Managing Resources
Setting Up Second-Level Disk Quota Parameters
Parallels Server Bare Metal provides the standard Linux quota package for working inside
Containers:
# pctl exec 101 rpm -q quota
quota-3.03-1.1.parallels
This command shows that the quota package installed in the Container is built and shipped by
Parallels. Use the utilities from this package (as is prescribed in your Linux manual) to set secondlevel quotas for the given Container. For example:
# ssh ct101
root@ct101's password:
Last login: Sat Jul 5 00:37:07 2009 from 10.100.40.18
[root@ct101 root]# edquota root
Disk quotas for user root (uid 0):
Filesystem
blocks
soft
hard
inodes
soft
hard
/dev/vzfs
38216
50000
60000
45454
70000
70000
[root@ct101 root]# repquota -a
*** Report for user quotas on device /dev/vzfs
Block grace time: 00:00; Inode grace time: 00:00
Block limits
File limits
User
used
soft
hard grace
used soft hard grace
---------------------------------------------------------------------root
-38218
50000
60000
45453 70000 70000
[the rest of repquota output is skipped]
[root@ct101 root]# dd if=/dev/zero of=test
dd: writing to `test': Disk quota exceeded
23473+0 records in
23472+0 records out
[root@ct101 root]# repquota -a
*** Report for user quotas on device /dev/vzfs
Block grace time: 00:00; Inode grace time: 00:00
Block limits
File limits
User
used
soft
hard grace
used soft hard grace
---------------------------------------------------------------------root
+50001
50000
60000
none
45454 70000 70000
[the rest of repquota output is skipped]
The above example shows the session when the root user has the disk space quota set to the
hard limit of 60,000 1KB blocks and to the soft limit of 50,000 1KB blocks; both hard and soft limits
for the number of inodes are set to 70,000.
It is also possible to set the grace period separately for block limits and inodes limits with the help
of the /usr/sbin/setquota command. For more information on using the utilities from the
quota package, consult the system administration guide shipped with your Linux distribution or
online manual pages included in the package.
120
Managing Resources
Checking Quota Status
As the server administrator, you can check the quota status for any Container with the vzquota
stat and vzquota show commands. The first command reports the status from the kernel and
shall be used for running Containers. The second command reports the status from the quota file
(located at /var/vzquota/quota.<CT_ID>) and shall be used for stopped Containers. Both
commands have the same output format.
The session below shows a partial output of Container 101 quota statistics:
# vzquota stat 101 –t
resource
usage
softlimit
1k-blocks
38281
1000000
inodes
45703
90000
User/group quota: on,active
Ugids: loaded 34, total 34, limit 100
Ugid limit was exceeded: no
hardlimit
1100000
91000
grace
User/group grace times and quotafile flags:
type block_exp_time inode_exp_time dqi_flags
user
0h
group
0h
User/group objects:
ID
type
resource
0
user 1k-blocks
0
user
inodes
[the rest is skipped]
usage
38220
45453
softlimit
50000
70000
hardlimit
60000
70000
grace status
loaded
loaded
The first three lines of the output show the status of first-level disk quotas for the Container. The
rest of the output displays statistics for user/group quotas and has separate lines for each user and
group ID existing in the system.
If you do not need the second-level quota statistics, you can omit the –t switch from the vzquota
command line.
121
Managing Resources
Cleaning Up Containers
The first-level quota assigned to this or that Container essentially shows how much space may be
occupied by the Container private files, i.e. not by the OS or common applications files. The real OS
and application files reside in the /vz/template directory on the server and practically do not
add up to the Container quota (except for the symlinks to them located inside the Container and
occupying insignificant space).
However, there are situations when one and the same application or application update is installed
not as a template, but separately inside each and every Container. A good example of this is the
CPanel application with its robust auto-update features. If a certain version of CPanel is installed in
a number of Containers, and then an update is released, CPanel automatically updates itself in all
these Containers, thus creating a vast amount of identical files (not symlinks already) throughout the
Containers. These files tell dramatically on the Container quotas, which may be avoided by putting
all the identical files to the server template area and creating symlinks instead of real files inside the
affected Containers.
The problem like the one described above can be solved in two ways:
1
A special subarea is created inside the server template area - /vz/template/vc - for
housing the files identical among multiple Containers with the help of the vzcache utility.
2
If the application or application update installed directly into one or more Containers has a
corresponding application template or template update installed on the server, the real files
inside the Containers are replaced with symlinks to the template files on the server with the help
of the vzpkg link utility. This utility is used to create symlinks to application EZ templates.
122
Managing Resources
Moving Container Files to the Cache Area
We will illustrate the effect produced by vzcache by copying one and the same huge dummy file
into two Containers. First, let us learn the disk space occupied by the whole /vz partition and by
the two Containers - Container 101 and Container 102:
# df /vz
Filesystem
/dev/hda3
# pctl exec 101 df
Filesystem
vzfs
# pctl exec 102 df
Filesystem
vzfs
1K-blocks
13756796
Used Available Use% Mounted on
1348292 11622123 11% /vz
1K-blocks
1048576
Used Available Use% Mounted on
22311
1026265
3% /
1K-blocks
1048576
Used Available Use% Mounted on
22311
1026265
3% /
After that, we copy the dummy file, which is around 600 MB in size, to the root of these Containers:
# cp foo /vz/root/101
# cp foo /vz/root/102
Now check the disk space once again:
# df /vz
Filesystem
/dev/hda3
# pctl exec 101 df
Filesystem
vzfs
# pctl exec 102 df
Filesystem
vzfs
1K-blocks
13756796
Used Available Use% Mounted on
2569060 10401355 20% /vz
1K-blocks
1048576
Used Available Use% Mounted on
632430
416146 61% /
1K-blocks
1048576
Used Available Use% Mounted on
632430
416146 61% /
We see that around 600 MB has been added to the space occupied by each Container and,
consequently, around 1.2 GB has been added to the space used on the /vz partition. Now it's
time to resort to vzcache to get rid of identical files inside the Containers:
# vzcache -v 101 102
Processing VZFSv2 Container 101
VZFSv2 Container 101
78 regular files
Processing VZFSv2 Container 102
VZFSv2 Container 102
78 regular files
During the command execution, vzcache does the following:
•
Looks for identical files inside Container 101 and Container 102.
•
Creates the CT_UUID subdirectory (where CT_UUID denotes the Container unique identifier
and can be determined by viewing the UUID parameters in the Container configuration file)
within the server template area (/vz/template/vc by default) for each Container.
•
Moves the identical files to the created subdirectories in the server template area.
Let us now take the final look at the disk space usage:
# df /vz
Filesystem
/dev/hda3
# pctl exec 101 df
Filesystem
1K-blocks
13756796
Used Available Use% Mounted on
1953053 11017362 16% /vz
1K-blocks
Used Available Use% Mounted on
123
Managing Resources
vzfs
# pctl exec 102 df
Filesystem
vzfs
1048576
1K-blocks
1048576
15105
1033471
2% /
Used Available Use% Mounted on
15138
1033438
2% /
As you can see, both the server and the Containers have each gained more than 600 MB of disk
space. In real life, the disk space is gained by caching not one huge file in two Containers but a
number of identical files across many Containers.
The operation of the vzcache utility may be customized to a certain extent by using vzcache
command line switches (see the Parallels Command Line Reference Guide for details).
Associating Container Files With Application Templates
It may often happen that a security update should immediately be applied to a package installed as
a template on the server and added to a number of Containers hosted there. However, it takes
certain time to prepare a template update, so the server and/or Container administrators are not
inclined to wait for it and they install the original security update directly inside the Containers. As to
the template update, it becomes available a few days afterwards. In other cases, a Container
administrator might not know that there is a certain template installed on the server, so they install
the corresponding application directly inside their Container.
To eliminate cluttering up the Container disk space with application files that are present as part of
an application template on the server, the vzpkg link utility is used. When executed, this utility
links your Container to the application EZ templates installed on the server. Assuming that you
manually installed the openssl application inside Container 101 running Fedora 8, you can use
the following command to replace the openssl files inside this Container with symlinks to these
files in the /vz/template/fedora-core/8/x86/config/app/openssl directory on the
server:
# vzpkg link 101
124
Managing Resources
Detaching Containers From Caches
Whereas the vzcache utility helps effectively gain disk space both on the Parallels server and
within Containers, there may be situations when it is necessary to detach a Container from its
cache and copy the cached files back to the Container private area. A typical example of this is
migrating a Container to another Parallels server. The migration is not possible if there are links in
the Container private area pointing to the /vz/template/vzcaches directory on the server.
To copy the cached files back to the Container private area, the vzuncache utility is used:
# vzuncache 101 -a
[Optimization messages skipped...]
Container 101
53 magic symlinks to convert
Container 101 will be detached from the following caches:
Cache name
Size
dhcp0-84.sw.ru-2005030316237
607972K
Now Container 101 can safely be migrated to another Parallels server. Note that unlike vzcache,
the vzuncache utility shall be called for only one Container at a time. The -a switch tells the utility
to detach the Container from all the cache directories specified in its configuration file as the value
of the VZCACHE parameter.
Managing Network Accounting and Bandwidth
This section explains how to perform the following tasks in Parallels Server Bare Metal:
•
configuring network classes
•
viewing network traffic statistics
•
turning on and off network bandwidth management
•
configuring bandwidth limits
125
Managing Resources
Network Traffic Parameters
The table below summarizes the network traffic parameters that you can control in Parallels Server
Bare Metal.
Parameter
Description
traffic_shaping
If set to yes, traffic limitations for outgoing traffic are set for virtual
machines and Containers. The default is no.
bandwidth
This parameter lists all network adapters installed on the server and their
bandwidth.
totalrate
This parameter defines the bandwidth to allocate for each network class. It
is active if traffic shaping is turned on.
rate
If traffic shaping is turned on, this parameter specifies the bandwidth
guarantee for virtual machines and Containers.
ratebound
If this parameter is set to yes, the bandwidth guarantee (the global rate
parameter) is also the limit for the virtual machine or Container, and the
virtual machine or Container cannot borrow the bandwidth from the
totalrate bandwidth pool.
126
Managing Resources
Configuring Network Classes
Parallels Server Bare Metal allows you to track the inbound and outbound network traffic as well as
to shape the outgoing traffic for a virtual machine and Container. To provide the ability to
distinguish between domestic and international traffic, a concept of network classes is introduced.
It is important to fully understand this notion, because network classes IDs are used in the values of
some network traffic parameters. A network class is a range of IP addresses for which Parallels
Server Bare Metal counts and shapes the traffic.
Classes are specified in the /etc/vz/conf/networks_classes file. The file is in the ASCII
format, and all empty lines and lines starting with the # sign are ignored. Other lines have the
following format:
<class_id> <IP_address>/<prefix_length>
where <class_id> defines the network class ID, and the <IP_address>/<prefix_length>
pair defines the range of IP addresses for this class. There may be several lines for each class.
Classes 0 and 1 have special meanings:
•
Class 0 defines the IP address range for which no accounting is performed. Usually, it
corresponds to the server subnet (the server itself and its virtual machines and Containers).
Setting up class 0 is not required; however, its correct setup improves performance.
•
Class 1 is defined by Parallels Server Bare Metal to match any IP address. It must be always
present in the network classes definition file. Therefore, it is suggested not to change the default
line in the networks_classes file.
1 0.0.0.0/0
If your virtual machines and Containers are using IPv6 addresses, you can also add the
following line to this file:
1 ::/0
Other classes should be defined after class 1. They represent exceptions from the "matchingeverything" rule of class 1. The example below illustrates a possible configuration of the network
classes definition file containing rules for both IPv4 and IPv6 addresses:
# server networks
0 192.168.0.0/16
0 fe80::/64
# any IP address (all traffic)
1 0.0.0.0/0
1 ::/0
# class 2 – addresses for the "foreign" traffic
2 10.0.0.0/8
2 2001:db88::/64
#
#
1
1
inside "foreign" network there
is a hole belonging to "local" traffic
10.10.16.0/24
2001:db88:3333::/64
127
Managing Resources
In this example, IPv4 addresses in the range of 192.168.0.0 to 192.168.255.255 and IPv6
addresses in the range of fe80:: to fe80::ffff:ffff:ffff:ffff are treated as class 0
addresses and no accounting is done for the traffic from virtual machines and Containers destined
to these addresses.
Class 2 matches the following IP addresses:
•
IPv4 addresses from 10.0.0.0 to 10.255.255.255 with the exception of addresses in the
sub-range of 10.10.16.0 to 10.10.16.255, which are treated as class 1.
•
IPv6 addresses from 2001:db88:: to 2001:db88::ffff:ffff:ffff:ffff with the
exception of addresses in the sub-range of 2001:db88:3333:: to
2001:db88:3333::ffff:ffff:ffff:ffff, which are also treated as class 1.
All other IP addresses (both IPv4 and IPv6) belong to class 1.
Note: After editing the /etc/vz/conf/networks_classes file, execute either the
/etc/init.d/vz accrestart or service vz accrestart command for the changes made to
the file to take effect.
128
Managing Resources
Viewing Network Traffic Statistics
In Parallels Server Bare Metal, you can view the current network traffic statistics using the following
utilities:
•
vznetstat to view the statistics for Containers
•
pnetstat to view the statistics for virtual machines
Viewing network statistics for Containers
To view the network statistics for Containers, execute the vznetstat utility:
# vznetstat
CTID
101
101
102
102
Net.Class
1
2
0
1
Input(bytes)
2202448
0
0
0
Input(pkts)
19527
0
0
0
Output(bytes)
9081832
0
0
0
Output(pkts)
19584
0
0
0
In this example, the traffic statistics for two Containers with IDs 101 and 102 is shown. Keep in
mind that the vznetstat utility displays statistics only on Containers that were started at least
once. So, some of your Containers may be not listed in the output.
The command output displays the following information:
•
Around 2 MB of data were uploaded to Container 101 and about 9 MB were downloaded from
it.
•
All the traffic was exchanged with servers from class 1 networks.
•
Container 102 has not yet exchanged any traffic with any network.
You can also view the network traffic statistics for a particular Container:
# vznetstat -v 101
CTID Net.Class Input(bytes) Input(pkts)
101
1
2202448
19527
101
2
0
0
Output(bytes) Output(pkts)
9081832
19584
0
0
The command above displays the network statistics for Container 101.
Viewing network statistics for virtual machines
To view the network traffic statistics for virtual machines, use the pnetstat utility:
# pnetstat
UUID
Net.Class Input(bytes) Input(pkts) Output(bytes) Output(pkts)
5747d4a9-d4ed...
0
0
0
0
0
5747d4a9-d4ed...
1
0
0
0
0
The pnetstat output is the same as the one produced by the vznetstat utility, except for the
first output column that displays virtual machine IDs. If you do not remember which ID is assigned
to which virtual machine, run the pctl list -i command.
129
Managing Resources
Note: In the current version of Parallels Server Bare Metal, you cannot view network traffic statistics for
individual virtual machines.
130
Managing Resources
Turning On and Off Network Bandwidth Management
Traffic shaping (also known as network bandwidth management) allows you to control what
network bandwidth a virtual machine or Container receives for outgoing traffic. Traffic shaping is off
by default in Parallels Server Bare Metal and is controlled by the TRAFFIC_SHAPING parameter in
the /etc/vz/vz.conf global configuration file.
Note: Incoming traffic cannot be controlled for virtual machines and Containers in Parallels Server Bare
Metal.
To turn traffic shaping on, you need to complete the following steps:
1
Set the value of TRAFFIC_SHAPING to yes in the global configuration file.
2
Correctly set up the BANDWIDTH and TOTALRATE parameters values.
3
Start traffic shaping with the /etc/init.d/vz shaperon command.
The BANDWIDTH variable is used for specifying the network rate, in kilobits per second, of available
network adapters. By default, it is set to eth0:100000, which corresponds to a 100Mb/s Fast
Ethernet card. If your server has more network adapters installed, update this parameter by listing
all the adapters participating in shaping. For example, if you have two Fast Ethernet cards, set the
parameter to eth0:100000 eth1:100000.
The TOTALRATE variable specifies the size of the so-called bandwidth pool for each network class
being shaped. The bandwidth from the pool can be borrowed by virtual machines and Containers
when they need more bandwidth for communicating with hosts from the corresponding network
class. It is used to limit the total available outgoing traffic virtual machines and Containers can
consume. The format of this variable is
<NIC>:<network_class>:<bandwidth_in_Kbits_per_second> and defines the pool
size per network class for a given network adapter. Multiple entries for different network classes
and adapters can be separated by spaces. The default value for TOTALRATE is eth0:1:4000,
which corresponds to the pool size of 4Mb/s for Network Class 1 on the first Ethernet adapter.
In the /etc/vz/vz.conf configuration file, you can also define the RATE variable whose value
amounts to the number of kilobits per second any virtual machine and Container is guaranteed to
receive for outgoing traffic with a network class on an Ethernet device. The default value of this
parameter is eth0:1:8, which means that any virtual machine and Container is guaranteed to
receive the bandwidth of at least 8 Kb/s for sending data to Class 1 hosts on the first Ethernet
device. This bandwidth is not the limit for a virtual machine and Container (unless the RATEBOUND
parameter is enabled for the virtual machine or Container); the virtual machine or Container can
take the needed bandwidth from the TOTALRATE bandwidth pool if it is not used by other virtual
machines and Containers.
After setting up the above parameters, start bandwidth management as follows:
# /etc/init.d/vz shaperon
Starting shaping: Ok
Set shaping on running Container :
131
Managing Resources
vz WARNING: Can't get tc class for Container(101).
vz WARNING: Can't access file /var/run/vz_tc_classes. \
Creating new one.
vz WARNING: Can't get tc class for Container(1).
Now you have activated the network bandwidth limits. To turn traffic shaping off temporarily, use
the /etc/init.d/vz shaperoff command. If you want to disable bandwidth management
permanently, set the TRAFFIC_SHAPING variable to no in the /etc/vz/vz.conf configuration
file.
132
Managing Resources
Configuring Network Bandwidth Management
The network bandwidth for outgoing traffic a virtual machine and Container receives is controlled by
two parameters: RATE and RATEBOUND.
Note: Incoming traffic cannot be controlled in the current version of Parallels Server Bare Metal.
The RATE parameter specifies the guaranteed outgoing traffic rate that a virtual machine or
Container receives. This rate can be specified differently for different network classes. Bandwidth
values are specified in Kb/s. It is recommended to increase this value in 8 Kb/s chunks and to set it
no lower than 8 Kb/s. The example below demonstrates how to set the RATE parameter for the
MyVM virtual machine and Container 101 to 16 Kb/s for network class 1 on the eth0 network
adapter:
# pctl set MyVM --rate 1:16
The VM has been successfully configured.
# pctl set 101 --rate eth0:1:16 --save
Saved parameters for Container 101
Note: For Containers, you can also configure the RATE parameter for different network adapters. For
virtual machines, you can set this parameter for the default network adapter (usually eth0) only. The
rates for all other network adapters can be configured in the /etc/vz/vz.conf global configuration
file.
The RATEBOUND parameter specifies whether the network bandwidth available to virtual machine
or Container for outgoing traffic is limited by the bandwidth specified in the RATE variable. By
default, this parameter is turned off for all newly created virtual machines and Containers. That
means that virtual machines and Containers are allowed to take free bandwidth from the
TOTALRATE pool. You can turn the RATEBOUND parameter on by using the --ratebound option
of the pctl set command, for example:
# pctl set MyVM --ratebound on
The VM has been successfully configured.
# pctl set 101 --ratebound yes --save
Saved parameters for Container 101
The actual network bandwidth available to virtual machines and Containers depends on the
number of virtual machines and Containers and the total sum of the RATE values, and normally
does not coincide with the bandwidth specified in their own RATE parameters. If the RATEBOUND
parameter is turned on, the virtual machine or Container bandwidth is limited by the value of the
RATE parameter.
If the the RATE and RATEBOUND parameters are not set for individual virtual machines and
Containers, the values from the /etc/vz/vz.conf configuration file are taken. By default,
Parallels Server Bare Metal does not set RATEBOUND, which corresponds to no, and RATE is set to
eth0:1:8.
133
Managing Resources
The network bandwidth management in Parallels Server Bare Metal works in the following way. The
bandwidth pool for a given network class (configurable through the TOTALRATE variable in the
global configuration file) is divided among the virtual machines and Containers transmitting data
proportionally to their RATE settings. If the total value of the RATE variables of all virtual machines
and Containers transmitting data does not exceed the TOTALRATE value, each virtual machine or
Container gets the bandwidth equal or greater than its RATE value (unless the RATEBOUND variable
is enabled for this virtual machine or Container). If the total value of the RATE variables of all virtual
machines and Containers transmitting data exceeds the TOTALRATE value, each virtual machine or
Container may get less than its RATE value.
The example below illustrates the scenario when Containers 101 and 102 have RATEBOUND set to
no, and the MyVM virtual machine has RATEBOUND set to yes. With the default TOTALRATE of
4096 Kb/s and RATE of 8 Kb/s, the bandwidth pool will be distributed according to the following
table:
Container 101
Container 102
MyVM
Consumed bandwidth
transmits
idle
idle
Container 101: 4096 Kb/s
idle
idle
transmits
MyVM: 8 Kb/s
transmits
transmits
idle
Container 101: 2048 Kb/s
Container 102: 2048 Kb/s
transmits
idle
transmits
Container 101: 4032 Kb/s
MyVM: 8 Kb/s
transmits
transmits
transmits
Container 101: 2016 Kb/s
Container 102: 2016 Kb/s
Container 103: 8 Kb/s
Once you configure the bandwidth settings, activate your changes by running the following
command:
# /etc/init.d/vz shaperrestart
Stopping shaping: Ok
Starting shaping: Ok
Set shaping on running Container: Ok
This command clears off all existing shaping settings and sets them again using the configuration
files of running virtual machines and Containers.
Managing Disk I/O Parameters
This section explains how to manage disk input and output (I/O) parameters in Parallels Server Bare
Metal systems.
134
Managing Resources
Configuring Priority Levels for Virtual Machines and Containers
In Parallels Server Bare Metal, you can configure the disk I/O (input/output) priority level of virtual
machines and Containers. The higher the I/O priority level, the more time the virtual machine or
Container will get for its disk I/O activities as compared to the other virtual machines and
Containers on the server. By default, any virtual machine and Container on the server has the I/O
priority level set to 4. However, you can change the current I/O priority level in the range from 0 to 7
using the --ioprio option of the pctl set command. For example, you can issue the following
command to set the I/O priority of Container 101 and the MyVM virtual machine to 6:
# pctl set 101 --ioprio 6 --save
Saved parameters for Container 101
# pctl set MyVM --ioprio 6
set ioprio 6
The VM has been successfully configured.
To check the I/O priority level currently applied to Container 101 and the MyVM virtual machine, you
can execute the following commands:
•
For Container 101:
# grep IOPRIO /etc/vz/conf/101.conf
IOPRIO="6"
•
For the MyVM virtual machine:
# pctl list --info | grep ioprio
cpu 2 VT-x accl=high mode=32 cpuunits=1000 ioprio=6 iolimit=0
135
Managing Resources
Configuring the Disk I/O Bandwidth
In Parallels Server Bare Metal, you can configure the bandwidth virtual machines and Containers
are allowed to use for their disk input and output (I/O) operations. Limiting the disk I/O bandwidth
can help you prevent the situations when high disk activities in one virtual machine or Container
(generated, for example, by transferring huge amounts of data to/from the virtual machine or
Container) can slow down the performance of other virtual machines and Containers on the
Parallels server.
By default, the I/O bandwidth limit for all newly created virtual machines and Containers is set to 0,
which means that no limits are applied to any virtual machines and Containers. To limit the disk I/O
bandwidth for a virtual machine or Container, you can use the --iolimit option of the pctl
set command. For example, the following command sets the I/O bandwidth limit for the MyVM
virtual machine to 10 megabytes per second (MB/s):
# pctl set MyVM --iolimit 10
Set up iolimit: 10
The VM has been successfully configured.
To set the limit for a Container, just specify its ID instead of the virtual machine name and the -save option to save the changes to the Container configuration file, for example:
# pctl set 101 --iolimit 10 --save
Set up iolimit: 10485760
Saved parameters for Container 101
By default, the limit is set in megabytes per second. However, you can use the following suffixes to
use other measurement units:
•
G: sets the limit in gigabytes per second (1G).
•
K: sets the limit in kilobytes per second (10K).
•
B: sets the limit in bytes per second (10B).
Note: In the current version of Parallels Containers, the maximum I/O bandwidth limit you can set for a
virtual machine and Container is 2 GB per second.
To check that the I/O speed limit has been successfully applied to the MyVM virtual machine and
Container 101, use the pctl list command:
# pctl list MyVM -o iolimit
10
# pctl list 101 -o iolimit
IOLIMIT
10485760
At any time, you can remove the I/O bandwidth limit set for the MyVM virtual machine and Container
101 by running these commands:
# pctl
Set up
The VM
# pctl
Set up
136
set MyVM --iolimit 0
iolimit: 0
has been successfully configured.
set 101 --iolimit 0 --save
iolimit: 0
Managing Resources
Saved parameters for Container 101
137
Managing Resources
Viewing Disk I/O Statistics
In Parallels Containers 4.6, you can view disk input and output (I/O) statistics for virtual machines
and Containers. To display the I/O statistics for all running virtual machines and Containers on the
physical server, you can run the pstat utility with the -a option. For example:
# pstat -a
7:18pm, up 1 day, 1:29, 2 users, load average: 0.00, 0.01, 0.00
CTNum 1, procs 127: R
2, S 125, D
0, Z
0, T
0, X
0
CPU [ OK ]: CTs
0%, CT0
1%, user
0%, sys
2%, idle 98%, lat(ms) 12/0
Mem [ OK ]: total 1560MB, free 627MB/402MB (low/high), lat(ms) 0/0
ZONE0 (DMA): size 16MB, act 0MB, inact 0MB, free 11MB (0/0/0)
ZONE1 (Normal): size 880MB, act 76MB, inact 104MB, free 616MB (3/4/5)
ZONE2 (HighMem): size 684MB, act 116MB, inact 153MB, free 402MB (0/1/1)
Mem lat (ms): A0 0, K0 0, U0 0, K1 0, U1 0
Slab pages: 65MB/65MB (ino 43MB, de 9MB, bh 2MB, pb 0MB)
Swap [ OK ]: tot 2502MB, free 2502MB, in 0.000MB/s, out 0.000MB/s
Net [ OK ]: tot: in 0.005MB/s
45pkt/s, out 0.000MB/s
1pkt/s
lo: in 0.000MB/s
0pkt/s, out 0.000MB/s
0pkt/s
eth0: in 0.005MB/s
45pkt/s, out 0.000MB/s
1pkt/s
sit0: in 0.000MB/s
0pkt/s, out 0.000MB/s
0pkt/s
Disks [ OK ]: in 0.000MB/s, out 0.000MB/s
CTID
ST
101
OK
2004838458 OK
IOUSED% IOWAIT% IOSPEED
IP
0.00
0.00
2/100MB/s 10.10.11.101
0.00
0.00 0.0/---KB/s 10.10.11.201
The information related to the virtual machines and Containers disk I/O statistics is at the end of the
command output. The table below explains the displayed I/O parameters:
Parameter
Description
IOUSED%
The percentage of time the disks are used by the virtual machine or
Container.
IOWAIT%
The percentage of time when at least one I/O transaction in the virtual
machine or Container is waiting for being served.
IOSPEED
The current speed of disk I/O operations in the virtual machine or Container
and the I/O limit set for this virtual machine or Container, if any. The value can
be displayed in bytes, kilobytes, megabytes, or gigabytes per second,
depending on the units you used to set the I/O limit.
The pstat -a command outputs the disk I/O statistics for all virtual machines and Containers that
are currently running on the physical server. In the example output above, this statistics is shown
for Container 101 and for the virtual machine with ID 2004838458. Note that the displayed ID is the
kernel virtual machine ID (that is, the ID the kernel on the physical server uses to refer to this virtual
machine) and is different from the unique identifier of the virtual machine. To view both identifiers of
a virtual machine, you can use the pctl list -i command, for example:
# pctl list -i MyVM
ID: {5c1fb1bb-4364-4b42-86b2-c584bdd2223b}
EnvID: 2075205468
Name: MyVM
...
138
Managing Resources
This output shows that the MyVM virtual machine has the unique identifier of 5c1fb1bb-4364-4b4286b2-c584bdd2223b and the kernel virtual machine ID of 2075205468.
Notes: For more information on pstat and its options, see the Parallels Server Bare Metal 5.0
Reference Guide.
Managing Memory Parameters for Containers
This section describes the VSwap memory management system introduced in Parallels Server Bare
Metal 5.0. You will learn to do the following:
•
Configure the main VSwap parameters for Containers (p. 140).
•
Set the memory allocation limit in Containers (p. 141).
•
Enhance the VSwap functionality (p. 142).
If you have upgraded to Parallels Server Bare Metal 5.0, you will also learn how the system
calculates the new VSwap values (p. 143) from the memory parameters that were applied to
Containers before the upgrade.
139
Managing Resources
Configuring Main VSwap Parameters
Parallels Server Bare Metal 5.0 introduces a new scheme for managing memory-related parameters
in Containers—VSwap. Like many other memory management schemes used on standalone Linux
computers, this scheme is based on two main parameters:
•
RAM. This parameter determines the total size of RAM that can be used by the processes of a
Container.
•
swap. This parameter determines the total size of swap that can be used by the Container for
swapping out memory once the RAM is exceeded.
Notes:
1. In Parallels Server Bare Metal 5.0, the new VSwap memory management scheme has replaced the
SLM scheme.
2. You can also set memory limits for and provide memory guarantees to Containers by configuring
multiple UBC (User Beancounter) parameters (numproc, numtcpsock, vmguarpages, and so on).
These parameters provide you with comprehensive facilities of customizing the memory resources in
respect of your Containers. However, this way of managing system resources is more complex and
requires more effort to be made on your part to adopt it to your system. For detailed information on UBC
parameters, refer to the Administrator's Guide to Managing UBC Resources available at
http://www.parallels.com/products/pvcl/resources/docs.
The new memory management scheme works as follows:
1
You set for a Container a certain amount of RAM and swap space that can be used by the
processes running in the Container.
2
When the Container exceeds the RAM limit set for it, the swapping process starts.
The swapping process for Containers slightly differs from that on a standalone computer. The
Container swap file is virtual and, if possible, resides in the Node RAM. In other words, when
the swap-out for a Container starts and the Node has enough RAM to keep the swap file, the
swap file is stored in the Node RAM rather than on the hard drive.
3
Once the Container exceeds its swap limit, the system invokes the OOM Killer for this
Container.
4
The OOM Killer chooses one or more processes running in the affected Container and forcibly
kills them.
By default, any newly created Container starts using the new memory management scheme. To
find out the amount of RAM and swap space set for a Container, you can check the values of the
PHYSPAGES and SWAPPAGES parameters in the Container configuration file, for example:
# grep PHYSPAGES /etc/vz/conf/101.conf
PHYSPAGES="65536:65536"
# grep SWAPPAGES /etc/vz/conf/101.conf
SWAPPAGES="65536"
140
Managing Resources
In this example, the value of the PHYSPAGES parameter for Container 101 is set to 65536. The
PHYSPAGES parameter displays the amount of RAM in 4-KB pages, so the total amount of RAM
set for Container 101 equals to 256 MB. The value of the SWAPPAGES parameter is also set to 256
MB.
To configure the amounts of RAM and swap space for Container 101, use the --ram and --swap
options of the vzctl set command. For example, you can execute the following command to
set the amount of RAM and SWAP in Container 101 to 1 GB and 512 MB, respectively:
# pctl set 101 --ram 1G --swap 512M --save
Saved parameters for Container 101
You can also use the --physpages and --swappages options to set the amount of RAM and
swap space for Containers. For more information on all VSwap options, consult the Parallels
Containers 4.6 Reference Guide.
Configuring the Memory Allocation Limit
When an application starts in a Container, it allocates a certain amount of memory for its needs.
Usually, the allocated memory is much more than the application actually requires for its execution.
This may lead to a situation when you cannot run an application in the Container even if it has
enough free RAM. To deal with such situations, the VSwap memory management scheme
introduces a new parameter—VM_Overcommit. Using this parameter, you can configure the
amount of memory applications in a Container may allocate, irrespective of the amount of RAM and
swap space assigned to the Container.
The amount of memory that can be allocated by applications of a Container is the sum of RAM and
swap space set for this Container multiplied by a memory overcommit factor. In the default (basic)
Container configuration file, this factor is set to 1.5. For example, if a Container is based on the
default configuration file and assigned 1 GB of RAM and 512 MB, the memory allocation limit for
the Container will be 2304 MB. You can configure this limit and set it, for example, to 3 GB by
running this command:
# vzctl set 101 --vm_overcommit 2 --save
Saved parameters for Container 101
This command uses the factor of 2 to increase the memory allocation limit to 3 GB:
(1 GB of RAM + 512 MB of swap) * 2 = 3 GB
Now applications in Container 101 can allocate up to 3 GB of memory, if necessary.
Note: For more information on Container configuration files, see Managing Container Resources
Configurations (p. 149).
141
Managing Resources
Tuning VSwap
The new management scheme can be extended by using UBC parameters. For example, you can
set the numfile parameter to configure the maximal number of processes and threads a
Container may create or the numproc parameter to specify the number of files that may be
opened by all processes in the Container. For detailed information on using UBC parameters,
consult the Administrator's Guide to Managing UBC Resources.
142
Managing Resources
Configuring Legacy Containers
If you upgraded from an earlier version of Parallels Server Bare Metal 5.0, all Containers
automatically start using the new memory management scheme after the upgrade. The system
automatically calculates the values of RAM and swap from the memory parameters that were
applied to a Container. These calculations are performed as follows:
SLM
If a Container uses only SLM parameters:
•
The amount of RAM is set to the value of the slmmemorylimit parameter.
•
The amount of swap is set to 0.
•
The memory allocation limit is set to the value of the slmmemorylimit parameter multiplied
by 1.5.
For example, if the slmmemorylimit parameter for Container 101 is set to 1 GB, then the
Container will have them set to the following values after the upgrade: RAM = 1 GB, swap = 0,
memory allocation limit = 1.5 GB.
UBC
If a Container uses only UBC parameters:
•
The amount of RAM is set to the soft limit of the privvmpages parameter.
•
The amount of swap is set to 0.
•
The memory allocation limit is set to the hard limit of the privvmpages parameter
For example, if the soft limit of privvmpages for Container 101 is set to 65536 pages and the
hard limit to 131072, then the Container will have the following parameters: RAM = 256 MB, swap
= 0, memory allocation limit = 2.
SLM and UBC
If a Container uses both SLM and UBC parameters:
•
The amount of RAM is set to the value of the slmmemorylimit parameter.
•
The amount of swap is set to 0.
•
The memory allocation limit is set to the value of the slmmemorylimit parameter multiplied
by 1.5.
For example, if the slmmemorylimit parameter for Container 101 is set to 1 GB, then the
Container will have them set to the following values after the upgrade: RAM = 1 GB, swap = 0,
memory allocation limit = 1.5 GB.
143
Managing Resources
Managing Memory Resources for Virtual
Machines
This section describes the process of managing memory resources for virtual machines.
Configuring Main Memory Resources
The main memory resources for virtual machines include the following:
•
main memory
•
video memory
The process of managing these resources is described below.
Configuring Main Memory
To configure the amount of memory that will be available to the virtual machine, use the -memsize option of the pctl set command. The following session shows how to change the
amount of memory for the MyVM virtual machine from 512 MB to 756 MB and to check that the
new value has been successfully set:
# pctl list -i MyVM | grep memory
memory 512Mb
# pctl set MyVM --memsize 756
Set the memsize parameter to 756Mb
The VM has been successfully configured.
# pctl list -i MyVM | grep memory
memory 756Mb
You can configure the memory size for both running and stopped virtual machines.
Configuring Video Memory
To set the amount of video memory to be available to the virtual machine's video card, use the -videosize option of the pctl set command. Assuming that the current video memory size of
the MyVM virtual machine is set to 32 MB, you can increase it to 64 MB by running the following
command:
# pctl set MyVM --videosize 64
Set the --videosize parameter to 64Mb.
The VM has been successfully configured.
To check that the new value has been successfully set, use this command:
# pctl list -i MyVM | grep video
video 64Mb
144
Managing Resources
Configuring Additional Memory Parameters
Parallels Server Bare Metal 5.0 comes with an improved memory management scheme. This
scheme is driven by two parameters:
•
Reserved memory limit. The reserved memory limit defines the amount of memory on a
Parallels server that can be used by all virtual machines hosted on this server.
•
Memory quota. The memory quota controls the memory consumption by a particular virtual
machine. This parameter is composite and includes the guarantee, limit, priority, and ballooning
settings.
The sections below describe how to configure both parameters.
Configuring the Reserved Memory Limit
The reserved memory limit defines the amount of memory that can be consumed by all virtual
machines on a Parallels server. The remaining memory on the server is reserved for applications
that run on the server itself.
By default, the reserved memory limit is calculated automatically and depends on the amount of
memory installed on a Parallels server:
•
If the server has less than 3 GB of memory installed, this formula is used: "total RAM on the
server" multiplied by 0.7. So if the server has 2 GB of memory, the reserved limit is set to 1.4
GB.
•
If the server has more than 3 GB of memory installed, this formula is used: "total RAM on the
server" minus 1 GB. So if the server has 16 GB of memory, the reserved limit is set to 15 GB.
To configure the default reserved memory limit, you can use the --mem-limit option of the
prlsrvctl set command. For example, the following command reserves 14 GB of memory for
use by virtual machines:
# prlsrvctl set --mem-limit 14336
Set memory limit: 14336Mb
The Server has been successfully configured.
To revert to the default settings, use this command:
# prlsrvctl set --mem-limit auto
Set memory limit: auto
The Server has been successfully configured.
Configuring the Memory Quota
The memory quota allows you to control the memory consumption by a particular virtual machine.
The quota control parameters include the following:
145
Managing Resources
•
Guarantee. The amount of memory a virtual machine is guaranteed to get on demand. If the
virtual machine requests more memory than is guaranteed, the memory allocation may fail (for
example, if there is no free memory on the Parallels server at the moment). Moreover, if the
guaranteed amount of memory of all virtual machines running on the server plus their overhead
exceeds the reserved limit, you will not be able to start another virtual machine. By default, the
guaranteed memory is calculated on the basis of RAM and video memory assigned to a virtual
machine and is about a half of its total memory.
Note: The overhead of a virtual machine depends on its configuration. For example, the overhead of a
virtual machine that has 1024 MB of RAM, 2 CPUs, 256 MB of video memory, a network adapter, a
sound card, and a USB controller and runs on modern hardware does not usually exceed 35-40 MB. To
check the overhead of a running virtual machine, open the /proc/parallels/vm/VM_ID/meminfo
file and look for the value of the Unreclaimable parameter. Keep in mind, however, that this value
may change over time.
•
Limit. The maximum amount of memory a virtual machine is allowed to consume. The virtual
machine cannot exceed this limit even if the Parallels server has a lot of free memory and the
virtual machine requires this memory. By default, no limit is set for all newly created virtual
machines, and any virtual machine may consume all free memory on the server.
•
Priority. The priority (from 1 to 100) that defines which virtual machine will get memory first. The
higher the priority of a virtual machine, the more chances it has to get memory when the
Parallels server has insufficient memory resources. By default, the priority is set to 50.
•
Ballooning. The maximum amount of memory the balloon driver in a virtual machine may
allocate for its needs. Memory ballooning is a technique that allows your system to reclaim
memory from virtual machines. To do this, a special balloon driver is loaded into each running
virtual machine. When the system requires free memory but does not have any, it sends a
command to the balloon driver in the virtual machine to increase its size. The balloon driver
inflates by allocating the requested amount of memory in the virtual machine and then gives this
memory to the system.
By default, the balloon driver can allocate up to 60% of RAM set for a virtual machine. For
example, if the amount of RAM for a virtual machine is set to 2 GB, the balloon driver can
allocate the maximum of 1.2 GB of memory.
To configure these quota control parameters for a specific virtual machine, you can use the -memquota parameter of the pctl set command. For example, the following command sets for
the MyVM virtual machine the memory guarantee to 512 MB, the limit to 2 GB, the priority to 70,
and the ballooning limit to 50% of RAM:
# pctl set MyVM --memquota 512:2048:70:50
Disable the memquota auto calculation.
Set the memquota_min parameter to 512Mb.
Set the memquota_max parameter to 1024Mb.
Set the memquota_prio parameter to 70.
Set the memquota_maxballoon parameter to 50.
The VM has been successfully configured.
To check that all parameters have been successfully set, use this command:
# pctl list -i MyVM | grep memory_quota
memory_quota min=512Mb max=2048Mb priority=70 maxballoon=50%
146
Managing Resources
To revert to the default settings, run this command:
# pctl set MyVM --memquota auto
Enable the memquota auto calculation.
The VM has been successfully configured.
Possible Usage Scenarios
Below you can find three examples that demonstrate how the new management scheme can help
service providers optimize their business.
"Ghetto" (payable RAM)
You have a number of non-priority customers whose virtual machines use only a fraction of the
RAM assigned to them. The virtual machines are stored on different physical servers. To optimize
the resources usage, you decide to migrate all underutilized virtual machines to one "Ghetto"
server. You then set (a) the guarantee for the migrated virtual machines to the minimum values
recommended for the operating systems running in these VMs plus the VM overhead and (b) the
memory limit to their RAM values. In this scenario:
•
You host a large number of virtual machines on a single physical server.
•
You ensure that all virtual machines can get enough memory for non-intensive operations, but
do not guarantee the optimal performance of memory-intensive operations (the limit is equal to
the RAM).
•
You charge customers for the amount of RAM assigned to their virtual machines.
"What you pay is what you get" (payable guarantee and limit)
You have a number of customers whose virtual machines require a certain amount of memory all
the time. For these customers, you configure their virtual machines by setting the memory
guarantee to the requested amount plus the VM overhead. You also set the memory limit equal to
the guarantee. In this scenario:
•
You charge customers for the set memory guarantee.
•
Customers can get only the memory they pay for (the guarantee is equal to the limit).
•
You ensure that every virtual machine can get the required amount of memory defined by its
guarantee.
•
No virtual machine can affect the performance of the host and other virtual machines on this
host. To meet this requirement, you need to make sure that all virtual machines on the host are
configured for use in the "What you pay is what you get" scenario.
"Burstable memory" (payable guarantee)
You have a number of customers whose virtual machines consume small amounts of memory most
of the time. Sometimes, memory-intensive operations may also run in the virtual machines. For
these virtual machines, you set the memory guarantee to the values that are high enough to run
non-intensive memory operations and the memory limit to "unlimited". In this scenario:
•
You charge customers for the set memory guarantee.
147
Managing Resources
•
You ensure that all virtual machines can get enough memory for non-intensive operations, and
such operations are running in the affected virtual machines most of the time.
•
When virtual machines require more memory than is defined by their guarantee, they get free
memory available on the host (the limit is set to "unlimited"). If the host does not have enough
memory, the virtual machines start competing for free memory.
Enabling Memory Hotplug for Virtual Machines
If a guest operating system supports the memory hotplug functionality, you can enable this
functionality for the virtual machine. Once the memory hotplug functionality is turned on, you can
configure the amount of memory available to your virtual machines even if they are running.
Currently, the following systems come with the memory hotplug support:
Linux (both x86 and x64 versions)
•
CentOS 5.3 and higher
•
Red Hat Enterprise LInux 5.3 and higher
•
Fedora 13 and higher
•
SUSE Linux Enterprise Server 10 and higher
•
Ubuntu 10.04 and higher
Windows
•
x64 version of Windows Server 2008 R2 (Datacenter Edition)
•
x86 and x64 versions of Windows Server 2008 (Standard Edition)
•
x86 and x64 versions of Windows Server 2008 (Enterprise Edition)
•
x86 and x64 versions of Windows Server 2008 (Datacenter Edition)
•
x86 and x64 versions of Windows Server 2003 (Enterprise Edition)
By default, memory hotplug support is disabled for all newly created virtual machines. To enable
this functionality, you can use the --mem-hotplug option of the pctl set command. For
example, to enable the memory hotplug support in the MyVM virtual machine that runs one of the
supported operating systems, stop the MyVM virtual machine and run this command:
# pctl set MyVM --memory-hotplug on
set mem hotplug: 1
The VM has been successfully configured.
Once the functionality is enabled, you can configure the amount of memory for the MyVM virtual
machine even it is running.
To disable the memory hotplug support in the MyVM virtual machine, use this command:
# pctl set MyVM --mem-hotplug off
set mem hotplug: 0
The VM has been successfully configured.
The changes will come into effect on the next virtual machine start.
148
Managing Resources
Managing Container Resources Configuration
Any Container is configured by means of its own configuration file. You can manage Container
configurations in a number of ways:
1
Using configuration sample files shipped with Parallels Server Bare Metal. These files are used
when a new Container is being created (for details, see Creating and Configuring New
Containers). Currently, the following configuration sample files are provided:
• basic. Use it for creating standard Containers.
• confixx. Use it for creating Containers that are to run the Confixx control panel.
• slm.plesk. Use it for creating Containers with the Plesk control panel.
• slm.256MB. Use it for creating Containers with 256 MB of main memory.
• slm.512Mb. Use it for creating Containers with 512 MB of main memory.
• slm.1024Mb. Use it for creating Containers with 1024 MB of main memory.
• slm.2048Mb. Use it for creating Containers with 2048 MB of main memory.
Note: Configuration sample files cannot contain spaces in their names.
Any sample configuration file can also be applied to an existing Container. You would do this if,
for example, you want to upgrade or downgrade the overall resources configuration of a
particular Container:
# pctl set 101 --applyconfig basic --save
This command applies all the parameters from the ve-basic.conf-sample file to Container
101.
When you install Parallels Server Bare Metal on your server, the default Container samples are
put to the /etc/vz/conf directory. They have the following format: ve-<name>.confsample (for example, ve-basic.conf-sample). In this connection you should keep in mind
the following when working with Container samples:
• When you create a Container using the pctl create command utility and base it on
some Container sample, this sample is taken from the /etc/vz/conf directory.
2
Using specific utilities for preparing configuration files in their entirety. The tasks these utilities
perform are described in the following subsections of this section.
3
The direct creating and editing of the corresponding Container configuration file
(/etc/vz/conf/<CT_ID>.conf). This can be performed with the help of any text editor.
The instructions on how to edit Container configuration files directly are provided in the four
preceding sections. In this case you have to edit all the configuration parameters separately,
one by one.
149
Managing Resources
Splitting Server Into Equal Pieces
It is possible to create a Container configuration roughly representing a given fraction of the server.
If you want to create such a configuration that up to 20 fully loaded Containers would be able to be
simultaneously running on the given server, you can do it as follows:
# cd /etc/vz/conf
# vzsplit -n 20 -f mytest
Config /etc/vz/conf/ve-mytest.conf-sample was created
Notice that the configuration produced depends on the given server resources. Therefore, it is
important to validate the resulted configuration file before trying to use it, which is done with the
help of the vzcfgvalidate utility. For example:
# vzcfgvalidate ve-mytest.conf-sample
Recommendation: kmemsize.lim-kmemsize.bar should be > 253952 \
(currently, 126391)
Recommendation: dgramrcvbuf.bar should be > 132096 (currently, 93622)
The number of Containers you can run on the server is actually several times greater than the value
specified in the command line because Containers normally do not consume all the resources that
are guaranteed to them. To illustrate this idea, let us look at the Container created from the
configuration produced above:
# pctl create 101 --ostemplate redhat-el5-x86 --config mytest
Creating Container private area (redhat-el5-x86)
Container is mounted
Postcreate action done
Container is unmounted
Container private area created
Container registered successfully
# pctl set 101 --ipadd 192.168.1.101 --save
Saved parameters for Container 101
# pctl start 101
Starting Container ...
Container is mounted
...
# vzcalc 101
Resource
Current(%) Promised(%) Max(%)
Memory
0.53
1.90
6.44
As is seen, if Containers use all the resources guaranteed to them, then around 20 Containers can
be simultaneously running. However, taking into account the Promised column output, it is safe to
run 40-50 such Containers on this server.
Note: If you generate a Container configuration sample using the vzsplit command line utility, the
resulting Container sample is put to the /etc/vz/conf directory. This sample can then be used by
pctl create when creating a new Container on its basis.
150
Managing Resources
Scaling Container Configuration
Any configuration or configuration sample file can prove insufficient for your needs. You might have
an application which does not fit into existing configurations. The easiest way of producing a
Container configuration is to scale an existing one.
Scaling produces a “heavier” or “lighter” configuration in comparison with an existing one. All the
parameters of the existing configuration are multiplied by a given number. A heavier configuration is
produced with a factor greater than 1, and a lighter one – with a factor between 0 and 1.
Note: If you create a new sample on the basis of an existing sample using the vzcfgscale command
line utility, the resulting Container sample is put to the /etc/vz/conf directory. This sample can then
be used by pctl create when creating a new Container on its basis.
The session below shows how to produce a configuration sample 50% heavier than the basic
configuration shipped with Parallels Server Bare Metal:
# cd /etc/vz/conf
# vzcfgscale -a 1.5 -o ve-improved.conf-sample ve-basic.conf-sample
# vzcfgvalidate ve-improved.conf-sample
Recommendation: kmemsize.lim-kmemsize.bar should be > 245760 \
(currently, 221184)
Recommendation: dgramrcvbuf.bar should be > 132096 (currently, 98304)
Validation completed: success
Now improved can be used in the pctl create command for creating new Containers.
It is possible to use the same technique for scaling configurations of the existing Containers. Notice
that the output file cannot be the same as the file being scaled. You have to save the scaling results
into an intermediate file.
151
Managing Resources
Validating Container Configuration
The system resource control parameters have complex interdependencies. The violation of these
interdependencies can be catastrophic for the Container. In order to ensure that a Container does
not break them, it is important to validate the Container configuration file before creating Containers
on its basis.
The typical validation scenario is shown below:
# vzcfgvalidate /etc/vz/conf/101.conf
Error: kmemsize.bar should be > 1835008 (currently, 25000)
Recommendation: dgramrcvbuf.bar should be > 132096 (currently, 65536)
Recommendation: othersockbuf.bar should be > 132096 \
(currently, 122880)
# pctl set 101 --kmemsize 2211840:2359296 --save
Saved parameters for Container 101
# vzcfgvalidate /etc/vz/conf/101.conf
Recommendation: kmemsize.lim-kmemsize.bar should be > 163840 \
(currently, 147456)
Recommendation: dgramrcvbuf.bar should be > 132096 (currently, 65536)
Recommendation: othersockbuf.bar should be > 132096 \
(currently, 122880)
Validation completed: success
The utility checks constraints on the resource management parameters and displays all the
constraint violations found. There can be three levels of violation severity:
Recommendation This is a suggestion, which is not critical for Container or server
operations. The configuration is valid in general; however, if the system
has enough memory, it is better to increase the settings as advised.
Warning
A constraint is not satisfied, and the configuration is invalid. The
Container applications may not have optimal performance or may fail
in an ungraceful way.
Error
An important constraint is not satisfied, and the configuration is invalid.
The Container applications have increased chances to fail
unexpectedly, to be terminated, or to hang.
In the scenario above, the first run of the vzcfgvalidate utility found a critical error for the
kmemsize parameter value. After setting reasonable values for kmemsize, the resulting
configuration produced only recommendations, and the Container can be safely run with this
configuration.
152
Managing Resources
Applying New Configuration Sample to Container
Parallels Server Bare Metal allows you to change the configuration sample file a Container is based
on and, thus, to modify all the resources the Container may consume and/or allocate at once. For
example, if Container 101 is currently based on the basic configuration sample and you are
planning to run the Plesk application inside the Container, you may wish to apply the slm.plesk
sample to it instead of basic, which will automatically adjust the necessary Container resource
parameters for running the Plesk application inside Container 101. To do this, you can execute the
following command on the server:
# pctl set 101 --applyconfig slm.plesk --save
Saved parameters for Container 101
This command reads the resource parameters from the ve-slm.plesk.conf-sample file
located in the /etc/vz/conf directory and applies them one by one to Container 101.
When applying new configuration samples to Containers, keep in mind the following:
•
All Container sample files are located in the /etc/vz/conf directory on the server and are
named according to the following pattern: ve-<name>.conf-sample. You should specify
only the <name> part of the corresponding sample name after the --applyconfig option
(slm.plesk in the example above).
•
The --applyconfig option applies all the parameters from the specified sample file to the
given Container, except for the OSTEMPLATE, TEMPLATES, VE_ROOT, VE_PRIVATE,
HOSTNAME, IP_ADDRESS, TEMPLATE, NETIF parameters (if they exist in the sample file).
•
You may need to restart your Container depending on the fact whether the changes for the
selected parameters can be set on the fly or not. If some parameters could not be configured
on the fly, you will be presented with the corresponding message informing you of this fact.
153
CHAPTER 5
Managing Services and Processes
This chapter provides information on what services and processes are, how they influence the
operation and performance of your system, and what tasks they perform in the system.
You will learn how to use the command line utilities in order to manage services and processes in
Parallels Server Bare Metal. In particular, you will learn how to monitor active processes in your
system, change the mode of the xinetd-dependent services, identify the Container ID where a
process is running by the process ID, start, stop, or restart services and processes, and edit the
service run levels.
Note: In the current version of Parallels Server Bare Metal, you cannot manage services and processes
in virtual machines using Parallels Server Bare Metal utilities. However, you can log in to a particular virtual
machine (e.g. via RDP to a Windows virtual machine and SSH to a Linux virtual machine) and manage its
services and processes in the same way you would manage them on a standalone computer.
In This Chapter
What Are Services and Processes ........................................................................... 155
Main Operations on Services and Processes ........................................................... 156
Managing Processes and Services .......................................................................... 157
Managing Services and Processes
What Are Services and Processes
Instances of any programs currently running in the system are referred to as processes. A process
can be regarded as the virtual address space and the control information necessary for the
execution of a program. A typical example of a process is the vi application running on your server
or inside your Linux-based Containers. Along with common processes, there are a great number of
processes that provide an interface for other processes to call. They are called services. In many
cases, services act as the brains behind many crucial system processes. They typically spend most
of their time waiting for an event to occur or for a period when they are scheduled to perform some
task. Many services provide the possibility for other servers on the network to connect to the given
one via various network protocols. For example, the nfs service provides the NFS server
functionality allowing file sharing in TCP/IP networks.
You may also come across the term "daemon" that is widely used in connection with processes
and services. This term refers to a software program used for performing a specific function on the
server system and is usually used as a synonym for "service". It can be easily identified by "d" at the
end of its name. For example, httpd (short for the HTTP daemon) represents a software program
that runs in the background of your system and waits for incoming requests to a web server. The
daemon answers the requests automatically and serves the hypertext and multimedia documents
over the Internet using HTTP.
When working with services, you should keep in mind the following. During the lifetime of a service,
it uses many system resources. It uses the CPUs in the system to run its instructions and the
system's physical memory to hold itself and its data. It opens and uses files within the file systems
and may directly or indirectly use certain physical devices in the system. Therefore, in order not to
decrease your system performance, you should run only those services on the Parallels server that
are really needed at the moment.
Besides, you should always remember that running services in the Host OS is much more
dangerous than running them in virtual machines and Containers. In case violators get access to
one of the virtual machines and Containers through any running service, they will be able to
damage only the virtual machine and Container where this service is running, but not the other
virtual machines and Containers on your server. The Parallels server itself will also remain unhurt.
And if the service were running on the Parallels server, it would damage both the server and all
virtual machines and Containers residing on it. Thus, you should make sure that you run only those
services on the server that are really necessary for its proper functioning. Launch all additional
services you need at the moment inside separate virtual machines and Containers. It can
significantly improve your system safety.
155
Managing Services and Processes
Main Operations on Services and Processes
The ability to monitor and control processes and services in your system is essential because of the
profound influence they have on the operation and performance of your whole system. The more
you know about what each process or service is up to, the easier it will be to pinpoint and solve
problems when they creep in.
The most common tasks associated with managing services running on the Parallels server or
inside a virtual machine and Container are starting, stopping, enabling, and disabling a service. For
example, you might need to start a service in order to use certain server-based applications, or you
might need to stop or pause a service in order to perform testing or to troubleshoot a problem.
For xinetd-dependent services, you do not start and stop but enable and disable services. The
services enabled in this way are started and stopped on the basis of the corresponding state of the
xinetd daemon. Disabled services are not started whatever the xinetd state.
In Parallels Server Bare Metal, you can manage services on the Parallels server and inside
Containers by means of special Linux command-line utilities. You can do it either locally or from any
server connected on the network.
As for processes, such Parallels Server Bare Metal utilities as vzps, vztop, vzpid enable you to
see what a process is doing and to control it. Sometimes, your system may experience problems
such as slowness or instability, and using these utilities can help you improve your ability to track
down the causes. It goes without saying that in Parallels Server Bare Metal you can perform all
those operations on processes you can do in a normal system, for example, kill a process by
sending a terminate signal to it.
156
Managing Services and Processes
Managing Processes and Services
In Parallels Server Bare Metal, services and processes can be managed using the following
Parallels command line utilities:
•
vzps
•
vzpid
•
vztop
•
vzsetxinetd.
With their help, you can perform the following tasks:
•
print the information about active processes on your Parallels server
•
view the processes activity in real time
•
change the mode of the services that can be either xinetd-dependent or standalone
•
identify the Container ID where a process is running by the process ID
Note: In the current version of Parallels Server Bare Metal, you cannot use Parallels Server Bare Metal
utilities for managing services and processes in virtual machines. However, you can log in to a particular
virtual machine (e.g. via RDP to a Windows virtual machine and SSH to a Linux virtual machine) and
manage its services and processes in the same way you would manage them on a standalone computer.
157
Managing Services and Processes
Viewing Active Processes and Services
The vzps utility provides certain additional functionality related to monitoring separate Containers
running on the Parallels server. For example, you can use the -E switch with the vzps utility to:
•
display the Container IDs where the processes are running
•
view the processes running inside a particular Container
vzps prints the information about active processes on your Parallels server. When run without any
options, vzps lists only those processes that are running on the current terminal. Below is an
example output of the vzps run:
$ vzps
PID TTY
4684 pts/1
27107 pts/1
TIME CMD
00:00:00 bash
00:00:00 vzps
Currently, the only processes assigned to the user/terminal are the bash shell and the vzps
command itself. In the output, the PID (Process ID), TTY, TIME, and CMD fields are contained. TTY
denotes which terminal the process is running on, TIME shows how much CPU time the process
has used, and CMD is the name of the command that started the process.
Note: The IDs of the processes running inside Containers and displayed by running the vzps command
on the Parallels server does not coincide with the IDs of the same processes shown by running the ps
command inside these Containers.
As you can see, the standard vzps command just lists the basics. To get more details about the
processes running on your server, you will need to pass some command line arguments to vzps.
For example, using the aux arguments with this command displays processes started by other
users (a), processes with no terminal or one different from yours (x), the user who started the
process and when it began (u). Besides, you can pass vzps the -E switch to sort the processes
by the Container IDs where they are running.
# vzps aux -E
USER PID %CPU %MEM
root
1 0.0 0.0
root
5 0.0 0.0
root
6 0.0 0.0
#27
7 0.0 0.0
root
9 0.0 0.0
root 1574 0.0 0.1
VSZ
1516
0
0
0
0
218
RSS
128
0
0
0
0
140
TTY
?
?
?
?
?
pts/4
STAT START
S
Jul14
S
Jul14
S
Jul14
S
Jul14
S
Jul14
S
09:30
TIME
0:37
0:03
3:20
0:00
0:00
0:00
COMMAND
init
[ubstatd]
[kswapd]
[bdflush]
[kinoded]
-bash
There is a lot more information now. The fields USER, %CPU, %MEM, VSZ, RSS, STAT, and
START have been added. Let us take a quick look at what they tell us.
The USER field shows you which user initiated the command. Many processes begin at system
start time and often list root or some system account as the USER. Other processes are, of course,
run by individuals.
158
Managing Services and Processes
The %CPU, %MEM, VSZ, and RSS fields all deal with system resources. First, you can see what
percentage of the CPU the process is currently utilizing. Along with CPU utilization, you can see the
current memory utilization and its VSZ (virtual memory size) and RSS (resident set size). VSZ is the
amount of memory the program would take up if it were all in memory; RSS is the actual amount
currently in memory. Knowing how much a process is currently eating will help determine if it is
acting normally or has spun out of control.
You will notice a question mark in most of the TTY fields in the vzps aux output. This is because
most of these programs were started at boot time and/or by initialization scripts. The controlling
terminal does not exist for these processes; thus, the question mark. On the other hand, the bash
command has a TTY value of pts/4. This is a command being run from a remote connection and
has a terminal associated with it. This information is helpful for you when you have more than one
connection open to the machine and want to determine which window a command is running in.
STAT shows the current status of a process. In our example, many are sleeping, indicated by an S
in the STAT field. This simply means that they are waiting for something. It could be user input or
the availability of system resources. The other most common status is R, meaning that it is currently
running.
Note: For detailed information on all vzps parameters, output fields, states of processes, etc., please
consult the vzps manual pages.
You can also use the vzps command to view the processes inside any running Container. The
example below shows you how to display all active processes inside Container 101:
# vzps -E 101
CTID
PID TTY
101 27173 ?
101 27545 ?
101 27555 ?
101 27565 ?
101 27576 ?
101 27583 ?
101 27584 ?
101 27587 ?
101 27596 ?
TIME CMD
00:00:01 init
00:00:00 syslogd
00:00:00 sshd
00:00:00 xinetd
00:00:03 httpd
00:00:00 httpd
00:00:00 httpd
00:00:00 crond
00:00:00 saslauthd
159
Managing Services and Processes
Monitoring Processes in Real Time
The vztop utility is rather similar to vzps but is usually started full-screen and updates
continuously with process information. This can help with programs that may infrequently cause
problems and can be hard to see with vzps. Overall system information is also presented, which
makes a nice place to start looking for problems.
The vztop utility can be run on the server just as the standard Linux top utility. The only features
that distinguish the vztop utility from top are the following:
•
vztop allows you to use the -E option that monitors only the processes belonging to the
Container whose processes you want to display.
•
You can use the e interactive command to temporarily view/hide the CTIDs where the
processes are running.
•
You can use the E interactive command to set the filter on the CTID field that helps you display
only the processes belonging to the given Container.
The vztop utility usually has an output like the following:
# vztop -E 101
17:54:03 up 20 days, 23:37, 4 users, load average: 2.13, 1.89, 1.75
305 processes: 299 sleeping, 3 running, 3 zombie, 0 stopped
CPU0 states: 20.1% user 51.2% system 0.0% nice 0.0% iowait 28.1% idle
CPU1 states: 21.2% user 50.0% system 0.0% nice 0.0% iowait 28.1% idle
Mem: 1031088k av, 969340k used, 61748k free, 0k shrd, 256516k buff
509264k active,
330948k inactive
Swap: 4056360k av,
17156k used, 4039204k free
192292k cached
CTID
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
101 27173 root 16 0
1616 604 1420 S 0.0 0.1 0:01.86 init
101 27545 root 16 0
1520 624 1356 S 0.0 0.1 0:00.34 syslogd
101 27555 root 25 0
4008 1700 3632 S 0.0 0.4 0:00.04 sshd
101 27565 root 25 0
2068 860 1740 S 0.0 0.2 0:00.05 xinetd
101 27576 root 16 0
7560 3180 6332 S 0.0 0.7 0:03.78 httpd
101 27587 root 16 0
2452 1036 1528 S 0.0 0.2 0:00.34 crond
101 27596 root 25 0
4048 1184 3704 S 0.0 0.2 0:00.01 saslauthd
As you can see, vztop provides an ongoing look at the processor activity in real time (the display
is updated every 5 seconds by default, but you can change that with the d command-line option or
the s interactive command). It displays a list of the most CPU-intensive tasks on the system and
can provide an interactive interface for manipulating processes. It can sort the tasks by CPU usage,
memory usage, and runtime. Specifying 101 after the -E option allows you to display only those
processes that are running inside Container 101 only. Besides, most features can be selected by
an interactive command, for example, the e and E commands described above.
Note: In the current version of Parallels Server Bare Metal, you cannot use the vztop utility for
monitoring processes in virtual machines.
160
Managing Services and Processes
Changing Services Mode
xinetd is a service used to start and stop a variety of data communication services. xinetd
starts on the Parallels server startup and waits for a connection request from a remote client that
wants to connect to the server. There can be a number of remote clients in the network, and each
of them can use different network protocols to establish connection to the server. In order not to
run all network services responsible for a specific protocol, which will negatively influence the
system performance, the system starts only the xinetd service. This service controls all other
network services and, at the connection time, it starts the corresponding service to process this
connection. In such a way, xinetd saves system resources allowing you to run only those
network services in the system that are really needed at the moment.
The vzsetxinetd utility allows you to switch Container services between the standalone and
xinetd mode. The services that can be either standalone or dependent on xinetd are
sendmail, sshd, proftpd, and courier-imap. Whereas they are xinetd-dependent by
default, in order to consume less resources, you may want to make them standalone due to the
following reasons:
•
The CPanel application does not recognize sshd if it is dependent on xinetd;
•
sendmail does not process some rules correctly if it is dependent on xinetd;
•
A number of control panel applications and some others are not able to manage xinetdbased services at all.
The courier-imapd, courier-imapds, courier-pop3d, and courier-pop3ds services
are provided by the courier-imap service, thus vzsetxinetd can manage these services via
the courier-imap service.
Let us assume that you wish to check the mode of the sendmail service and set it to standalone
if it is in the xinetd mode. First, you should check the current status of the sendmail service.
To this effect, type the following command in the command line:
# vzsetxinetd -s 222 sendmail
where 222 is the Container ID, sendmail denotes the name of the corresponding service, and the
-s option gets the status of the sendmail service of the Container with ID 222. The output will tell
you if this service has the standalone or xinetd mode:
sendmail is xinetd service
In our case it is in the xinetd mode. Now you can change the mode of the sendmail service to
standalone. To make it standalone, type the following line:
# vzsetxinetd 222 sendmail off
sendmail is standalone service
where off specifies that the sendmail service should be set to the standalone mode. The output
confirms that the sendmail service is now standalone.
161
Managing Services and Processes
For more information on the vzsetxinetd utility, consult the corresponding man pages or refer to
the Parallels Command Line Reference Guide.
Notes:
1. You cannot use the vzsetxinetd utility to change the mode of the xinetd-dependent services in
Containers where the Debian 3.0 OS template is installed.
2. In the current version of Parallels Server Bare Metal, you cannot use the vzsetxinetd utility for
managing services in virtual machines.
Determining Container Identifiers by Process IDs
Each process is identified by a unique PID (process identifier), which is the entry of that process in
the kernel's process table. For example, when you start Apache, it is assigned a process ID. This
PID is then used to monitor and control this program. The PID is always a positive integer. In
Parallels Server Bare Metal, you can use the vzpid (retrieve process ID) utility to print the
Container ID the process with the given id belongs to. Multiple process IDs can be specified as
arguments. In this case the utility will print the Container number for each of the processes.
The typical output of the vzpid utility is shown below:
# vzpid 12
Pid
VEID
12
101
Name
init
In our example the process with the identifier 12 has the name 'init' and is running in the Container
with ID 101.
Note: You can also display the Container ID where the corresponding process is running by using the
vzps utility.
Starting, Stopping, and Restarting Services
You can manage (i.e. start, stop, and restart) services by using the command line. For example,
you wish to start the httpd service. To do this, execute the following command:
[root@ct222 /]# service httpd start
where service is the standard Linux command, httpd denotes the name of the corresponding
service, and start is the command that will launch this service. In order to check that the httpd
service was successfully launched, you can either type the following Linux command:
[root@ct222 /]# service httpd status
or use the vzps utility when working on your server or the ps utility when working inside your
Containers and passing them the x argument. The output will tell you if the httpd service is
running in your system or not.
162
CHAPTER 6
Managing Parallels Server Bare Metal
Network
The given chapter familiarizes you with the Parallels Server Bare Metal network structure,
enumerates Parallels networking components, and explains how to manage these components in
your working environments. In particular, it provides the following information:
•
How you can manage network adapters on the Parallels server.
•
What Virtual Networks are and how you can manage them on the Parallels server.
•
How to create virtual network adapters inside your virtual machines and Containers and
configure their parameters.
•
How to connect virtual machines and Containers to different networks.
In This Chapter
Managing Network Adapters on the Parallels Server ................................................ 164
Networking Modes in Parallels Server Bare Metal..................................................... 166
Configuring Virtual Machines and Containers in Host-Routed Mode ......................... 173
Configuring Virtual Machines and Containers in Virtual Network Mode ...................... 174
Managing Private Networks ..................................................................................... 187
Managing Parallels Server Bare Metal Network
Managing Network Adapters on the Parallels
Server
Network adapters installed on the Parallels server are used to provide virtual machines and
Containers with access to each other and to external networks. During the installation, Parallels
Server Bare Metal registers all physical and VLAN network adapters available on the server. In
addition to that, it creates a number of VLAN adapters on the server. Once Parallels Server Bare
Metal has been successfully installed, you can perform the following operations on network
adapters:
•
List the adapters currently available on the server.
•
Create new VLAN adapters on the server.
Note: For more information on Virtual Networks, refer to Managing Virtual Networks (p. 174).
These operations are described in the following subsections in detail.
164
Managing Parallels Server Bare Metal Network
Listing Adapters
You can view the physical, virtual, and VLAN network adapters on your Parallels server using the
vznetcfg utility. For example, you can execute the following command to list the available
adapters:
# vznetcfg if list
Name
Type
Network ID
eth0
nic
Bridged
br1
bridge Bridged
br0
bridge Host-Only
Addresses
10.30.18.41/16,fe80::20c:29ff:fee8:9419/64,dhcp
fe80::200:ff:fe00:0/64
fe80::200:ff:fe00:0/64
The information on adapters is presented in the table with the following columns:
Column Name
Description
Name
Adapter name.
Type
Type of the network adapter. It can be one of the following:
•
nic denotes a physical adapter.
•
vlan stands for a VLAN adapter.
•
bridge is a virtual bridge automatically created for each Virtual
Network on the Parallels server.
•
vethX is a virtual network adapter automatically created for each
veth network adapter in each Container.
•
vmeN is a virtual network adapter automatically created for each
network adapter that exists in a virtual machine and operates in
the virtual network mode.
Network ID
ID of the Virtual Network to which the network adapter is connected. Detailed
information on Virtual Networks is provided in Managing Virtual Networks
(p. 174).
Addresses
IP address and subnet mask assigned to the network adapter. dhcp denotes
that the adapter gets its network parameters from a DHCP server.
165
Managing Parallels Server Bare Metal Network
Creating VLAN Adapters
Parallels Server Bare Metal allows you to create new VLAN adapters on the Parallels server. You
can use these adapters later on to connect your virtual machines and Containers to any of the
available Virtual Networks (for more information on Virtual Networks, see Managing Virtual
Networks (p. 174)). VLAN adapters can be made using the vznetcfg vlan add command. To
create a new VLAN adapter, you need to specify two parameters.
•
VLAN ID—an arbitrary integer number which will uniquely identify the virtual LAN among other
VLANs on the server.
•
Physical network adapter on the server to which the VLAN is to be bound.
For example, you can execute the following command to make a new VLAN adapter, associate it
with a VLAN having the ID of 5 (i.e. with VLAN 5), and attach the VLAN adapter to the eth0
physical adapter on the server:
# vznetcfg vlan add eth0 5
To check that the VLAN adapter has been successfully created, execute the following command:
# vznetcfg
Name
eth0
eth0.5
...
if list
Type
Network ID
nic
vlan
Addresses
192.168.0.150/22,dhcp
VLAN adapters can be easily identified by the vlan designation shown in the Type column of the
command output. As you can see, only one VLAN adapter currently exists on the server. It is
assigned the name of eth0.5. This name is generated automatically on the basis of the specified
VLAN ID and the name of the physical adapter to which the VLAN adapter is tied.
At any time, you can delete the eth0.5 VLAN adapter and thus destroy VLAN 5 by issuing the
following command:
# vznetcfg vlan del eth0.5
Networking Modes in Parallels Server Bare Metal
In Parallels Server Bare Metal, any virtual machine and Container can operate in one of the two
networking modes:
•
host-routed
•
virtual network
Detailed information on these modes is given in the following sections.
166
Managing Parallels Server Bare Metal Network
Host-Routed Mode
The implementations of host-routed mode for Containers and virtual machines have much in
common but also show some differences. Therefore, we describe them in different sections.
Host-routed mode for Containers
By default, a newly created Container starts operating in the host-routed mode. In this mode, the
Container is connected with the other Containers on the server and with the server itself using a
special virtual network adapter called venet0. The picture below provides an example of the
network structure when all Containers (Container #1, Container #2, Container #3) are functioning in
the host-routed mode.
167
Managing Parallels Server Bare Metal Network
All Containers use the venet0 virtual adapter as the default gateway to send and receive data
to/from other networks (shown as the PUBLIC NETWORK in the picture above). The procedure of
handling incoming and outgoing IP packets can be described as follows:
•
All IP packets from Containers come to the venet0 adapter and are redirected through a
public IP address of the server to the corresponding server on the public network.
•
All IP packets coming from external networks and destined for Container IP addresses reach
the public IP address of the server first and, afterwards, are sent through venet0 to the IP
addresses of the corresponding Containers.
The venet0 adapter is also used to exchange the traffic among Containers. All network traffic of a
Container is isolated from that of the other Containers—that is, all Containers are protected from
each other in the way that makes traffic snooping impossible.
Host-routed mode for virtual machines
By default, a new virtual machine is created with the network adapter that operates in the virtual
network mode. To change the mode to host-routed, you can use the pctl set command or
Parallels Management Console. In the host-routed mode, all virtual machines use special virtual
adapters on the server to connect to each other, the server, and computers on external networks.
The picture below demonstrates an example network structure when all virtual machines (VM #1,
VM #2, and VM #3) are operating in the host-routed mode.
168
Managing Parallels Server Bare Metal Network
In this network:
•
All virtual machines use special networks adapters (Virtual Adapter 1, Virtual Adapter 2, and
Virtual Adapter 3) as the default gateways to send and receive data to/from other networks.
•
All IP packets from virtual machines come to the corresponding adapters and are redirected
through a public IP address of the server to the destination computer.
•
All IP packets coming from external networks and destined for virtual machines reach the public
IP address of the first and, afterwards, are sent through virtual adapters to the IP addresses of
the destination virtual machine.
Virtual adapters are also used to exchange the traffic among virtual machines. All network traffic of
a virtual machine is isolated from that of the other virtual machines—that is, all virtual machines are
protected from each other in the way that makes traffic snooping impossible.
169
Managing Parallels Server Bare Metal Network
Virtual Network Mode
The implementations of virtual network mode for Containers and virtual machines have much in
common but also show some differences. Therefore, we describe them in different sections.
Virtual network mode for Containers
You can create veth virtual adapters in Containers and make them operate in the virtual network
mode. The following figure represents an example of the network structure where all Containers
(Container #1 and Container #2) are operating in the virtual network mode.
170
Managing Parallels Server Bare Metal Network
In the virtual network mode, a separate veth virtual adapter is created for each Container. Any
veth virtual adapter consists of two interfaces:
•
An Ethernet interface in the Container. This interface represents a counterpart of a physical
network adapter installed on a standalone server. As any other physical adapter, it has a MAC
address, can be assigned one or more IP addresses and included in different network
environments, and so on. Refer to the Configuring veth Adapter Parameters section (p. 181)
for detailed information on configuring Ethernet interfaces in Containers.
•
An Ethernet interface on the server. This interface is responsible for the adapter operation in the
server context and mostly used to maintain the interaction and communication between the
server and the Ethernet interface in the Container. Each Ethernet interface is assigned a MAC
address. Detailed information on managing Ethernet interfaces on the server is provided in the
Configuring veth Adapter Parameters section (p. 181).
Both interfaces are closely linked to each other, which means that an IP packet entering one
interface will always come out from the other one.
Virtual network mode for virtual machines
By default, a new virtual machine is created with the network adapter that operates in the host-only
mode. The figure below demonstrates an example network structure where all virtual machines are
operating in the virtual network mode.
171
Managing Parallels Server Bare Metal Network
In the virtual network mode, a separate vme virtual adapter is created for each Container. Any vme
virtual adapter consists of two interfaces:
•
An Ethernet interface in the Container. This interface represents a counterpart of a physical
network adapter installed on a standalone server. As any other physical adapter, it has a MAC
address, can be assigned one or more IP addresses and included in different network
environments, and so on. Refer to the Configuring veth Adapter Parameters section (p. 181)
for detailed information on configuring Ethernet interfaces in virtual machine.
•
An Ethernet interface on the server. This interface is responsible for the adapter operation in the
server context and mostly used to maintain the interaction and communication between the
server and the Ethernet interface in the virtual machine. Each Ethernet interface is assigned a
MAC address. Detailed information on managing Ethernet interfaces on the server is provided in
the Configuring veth Adapter Parameters section (p. 181).
Both interfaces are closely linked to each other, which means that an IP packet entering one
interface will always come out from the other one.
Differences Between Host-Routed and Virtual Network Modes
The virtual network mode demonstrates a number of differences as compared to the host-routed
mode:
•
Each veth or vme virtual adapter has a MAC address assigned to it while a host-routed
adapter does not have any. Thanks to this fact:
• Any virtual machine and Container can see all broadcast and multicast packets received
from or sent to the selected network adapter on the server.
• Using veth or vme virtual adapters in virtual machines and Containers allows you to host
DHCP or Samba servers in these virtual machines and Containers.
•
There is no more need to assign all network settings (IP addresses, subnet mask, gateway, and
so on) to virtual machines and Containers from the server. All network parameters can be set
from inside virtual machines and Containers.
•
veth and vme adapters can be bridged among themselves and with other devices. If several
veth and vme adapters are united into a bridge, this bridge can be used to handle network
traffic for the virtual machines and Containers whose veth and vme adapters are included in
the bridge.
•
Due to the fact that veth and vme adapters act as full members on the network (rather than
'hidden' beyond virtual networks adapters on the server), they are more prone to security
vulnerabilities: traffic sniffing, IP address collisions, and so on. Therefore, veth and vme
adapters are recommended for use in trusted network environments only.
172
Managing Parallels Server Bare Metal Network
Configuring Virtual Machines and Containers in
Host-Routed Mode
You can configure the following parameters of network adapters that operate in the host-routed
mode:
•
IP addresses and network masks
•
DNS servers
•
DNS search domains
Setting IP addresses
The session below how to set IP addresses for the MyVM virtual machine and Container 101
#
#
#
#
pctl
pctl
pctl
pctl
set
set
set
set
MyVM --device-set net0 --ipadd 10.0.186.100/24
MyVM --device-set net0 --ipadd 1fe80::20c:29ff:fe01:fb07
101 --ipadd 10.0.186.101/24 --save
101 --ipadd fe80::20c:29ff:fe01:fb08 --save
net0 in the commands above denotes the network card in the VM virtual machine to assign the IP
address to. You can view all network cards of a virtual machine using the pctl list VM_name
-i command. For Container 101, you do not need to specify the network card name; pctl set
automatically performs the operation on the default adapter that always operates in the host-routed
mode.
Setting DNS server addresses
To set a DNS server for the MyVM virtual machine and Container 101, you can use the following
commands.
# pctl set MyVM --device-set net0 --nameserver 192.168.1.165
# pctl set 101 --nameserver 192.168.1.165 --save
Setting DNS search domains
To set a DNS search domain for the MyVM virtual machine and Container 101, run these
commands:
# pctl set MyVM --device-set net0 --searchdomain 192.168.10.10
# pctl set 101 --searchdomain 192.168.10.10 --save
Notes:
1. You can configure the network settings only of virtual machines that have Parallels Tools installed.
2. Network adapters operating in the routed mode must have at least one static IP address assigned.
173
Managing Parallels Server Bare Metal Network
3. To assign network masks to Containers operating in the venet0 networking mode, you must set the
USE_VENET_MASK parameter in the /etc/vz/vz.conf configuration file to yes.
4. Containers can have only one network adapter operating in the host-routed mode. This adapter is
automatically created when you create a virtual machine.
Switching network adapters to the host-routed mode
By default, a virtual adapter in any newly created virtual machine starts operating in connected to
the virtual network mode (see the Connecting Virtual Machines to Virtual Networks (p. 186) for
details). To change the current network mode to host-routed, you can run the following command:
# pctl set VM_Name --device-set Net_ID --type routed
For example, to set the net0 adapter in the MyVM virtual machine to operate in the host-routed
mode, use this command:
# pctl set MyVM --device-set net0 --type routed
Configure net0 (+) dev='vme417795ba.0' type=routed mac=001C424BD617 card=e1000
state=disconnected
The VM has been successfully configured.
Configuring Virtual Machines and Containers in
Virtual Network Mode
This section describes all operations related to configuring virtual machines and Containers that
operate in virtual network mode.
Managing Virtual Networks
A virtual network acts as a binding interface between a virtual network adapter in a virtual machine
and Container and the corresponding network adapter on the Parallels server. Using virtual
networks, you can include virtual machines and Containers in different networks. Parallels Server
Bare Metal enables you to manage virtual networks as follows:
•
Create a new Virtual Network and remove an existing one.
•
Configure the parameters of an existing Virtual Network.
•
List the existing Virtual Networks.
•
Delete a Virtual Network that you do not need any more.
These operations are described in the following subsections in detail.
174
Managing Parallels Server Bare Metal Network
Creating a Virtual Network
Virtual networks serve as binding interfaces between the virtual network adapters in virtual
machines and Containers and the physical, VLAN, and virtual network adapters on the Parallels
server. Using virtual networks, you can connect virtual machines and Containers to different
networks.
By default, Parallels Server Bare Metal creates the following virtual networks on the server:
•
Bridged. This virtual network is connected to one of the physical adapters on the Parallels
server (as a rule, eth0) and provides virtual machines and Containers included in this virtual
network with access to the network behind this physical adapter.
•
Host-only. This virtual network is connected to a special virtual adapter on the server and allows
the virtual machines and Containers joined to this virtual network to access only the server and
the other virtual machines and Containers on this network.
You can also create your own virtual networks using the prlsrvctl or vznetcfg utility. For
example, to make a new virtual network with the name of vznetwork1, you can rung one of the
following commands:
# vznetcfg net new vznetwork1
or
# prlsrvctl net add vznetwork1
By default, both commands create host-only virtual networks. However, you can change their types
using the prlsrvctl utility; see Configuring Virtual Network Parameters (p. 176) for details.
In the current version of Parallels Server Bare Metal, you can create
•
Up to 16 host-only virtual networks.
•
One or more bridged virtual networks. The number of virtual networks depends on the number
of physical and VLAN adapters available on the Parallels server. One virtual network can be
connected to only one physical or VLAN adapter.
Viewing Bridges
A virtual network is associated with a bridge that is automatically made on the Parallels server when
you create the virtual network and serves as the basis for the virtual network operation. To find out
what bridge is associated with what virtual network, you can run the following command:
# vznetcfg if list
Name Type
Network ID
eth0 nic
Bridged
br1 bridge Bridged
br0 bridge Host-Only
Addresses
10.31.252.116/16,fe80::2a9:40ff:fe0f:b6f2/64,dhcp
fe80::200:ff:fe00:0/64
10.37.130.2/24,fdb2:2c26:f4e4::1/64,fe80::200:ff:fe00:0/64
The bridges existing on the Parallels server are listed in the Name column and can be easily
identified by the br prefix.
175
Managing Parallels Server Bare Metal Network
Note: Detailed information on the vznetcfg and prlsrvctl utilities is provided in the Parallels Server
Bare Metal 5.0 Command Line Reference Guide.
Configuring Virtual Network Parameters
Parallels Server Bare Metal allows you to configure the following parameters for a virtual network:
•
the name assigned to the virtual network
•
the networking mode in which the virtual network is operating
•
the description of the virtual network
All these operations can be performed using the prlsrvctl utility. Let us assume that you want
to configure the psbmnet1 virtual network. This virtual network is currently configured as a hostonly network and has the following description set: This is a host-only virtual
network. To change these parameters, you can execute the following command:
# prlsrvctl net set vznetwork1 -n psbm_network1 -t bridged --ifname eth1 -d "This is
now a bridged virtual network"
This command configured the psbmnet1 virtual network as follows:
1
Changes the virtual network name to psbm_network1.
2
Changes the virtual network type to bridged.
3
Changes the virtual network description to the following: This is a now bridged
virtual network.
For more information on the prlsrvctl utility, refer to the Parallels Server Bare Metal 5.0
Command Line Reference Guide.
176
Managing Parallels Server Bare Metal Network
Listing Virtual Networks
To list the virtual networks existing on the Parallels server, you can use either the vznetcfg or
prlsrvctl utility.
Listing virtual networks with vznetcfg
To list the virtual networks on your server using the vznetcfg utility, execute the following
command:
# vznetcfg net list
Network ID
Status
Host-Only
active
Bridged
active
psbmnet1
active
Master Interface
Slave Interfaces
eth0
eth1
In the example above, three virtual networks—psbmnet1 and two default virtual networks—exist
on the Parallels server. The information on these virtual networks is presented in the table with the
following columns:
Column Name
Description
Network ID
The ID assigned to the virtual network.
Status
Indicates the status of the virtual network. It can be one of the following:
•
active: the virtual network is up and running.
•
configured: the information on the virtual network is
present in the /etc/vz/vznet.conf file on the server, but
the bridge to which the virtual network is bound is down or
does not exist.
Note: Detailed information on the vznet.conf file is given in
the Parallels Server Bare Metal 5.0 Command Line Reference
Guide.
Master Interface
Displays the adapter on the server connected to the virtual network, if
any.
Slave Interfaces
Lists the adapters in virtual machines and Containers joined to the virtual
network, if any.
Listing virtual networks with prlsrvctl
You can also use the prlsrvctl utility to list the virtual networks existing on your server. To do
this, run the following command:
# prlsrvctl net list
Network ID
Type
Host-Only
host-only
Bridged
bridged
psbmnet1
bridged
Bound To
eth0
eth1
This utility displays the following information on virtual networks:
177
Managing Parallels Server Bare Metal Network
Column Name
Description
Network ID
The name assigned to the virtual network.
Type
The networking mode set for the virtual network.
Bound To
The adapter on the Parallels server connected to the virtual networks, if
any.
Connecting Virtual Networks to Adapters
By connecting an adapter on the physical server to a virtual network, you can join all virtual
machines and Containers included in the virtual network to the network to which the corresponding
adapter is connected.
Let us assume the following:
•
The eth1 physical adapter and the psbmnet1 virtual network exist on the Parallels server. For
information on creating virtual networks, see Creating a Virtual Network (p. 175).
•
The eth1 physical adapter is connected to the local network.
•
The MyVM virtual machine is connected to the psbmnet1 virtual network. Detailed information
on joining virtual machines and Containers to virtual networks is given in Connecting
Containers to Virtual Networks (p. 182) and Connecting Virtual Machines to Virtual
Networks (p. 186).
To connect the eth1 adapter to the psbmnet1 virtual network and thus to join the MyVM virtual
machine to the network behind eth1, run this command on the server:
# vznetcfg net addif psbmnet1 eth1
To check that the eth1 physical adapter has been successfully added to the psbmnet1 virtual
network, you can execute the following command:
# vznetcfg if list
Name
Type
Network ID
eth1
nic
psbmnet1
...
Addresses
10.31.252.116/16,fe80::2a9:40ff:fe0f:b6f2/64,dhcp
As you can see, the eth1 adapter is now joined to the psbmnet1 virtual network. That means that
the MyVM virtual machine whose virtual network adapter is connected to psbmnet1 can access
the local network behind eth1.
At any time, you can disconnect the eth1 physical adapter from the psbmnet1 virtual network
(and thus detach the MyVM virtual machine from the local network) by running the following
command:
# vznetcfg net delif eth1
178
Managing Parallels Server Bare Metal Network
Deleting Virtual Networks
At any time, you can remove a virtual network that you do not need any more from the physical
server. To do this, you can use both the vznetcfg and prlsrvctl utilities. For example, you
can delete the psbmnet1 virtual network by running one of the following commands:
# vznetcfg net del psbmnet1
or
# prlsrvctl net del psbmnet1
To check that psbmnet1 has been successfully removed, execute one of these commands:
# vznetcfg net list
Network ID
Status
Host-Only
active
Bridged
active
Master Interface
Slave Interfaces
eth0
or
# prlsrvctl net list
Network ID
Type
Host-Only
host-only
Bridged
bridged
Bound To
eth0
Note: Detailed information on the vznetcfg and prlsrvctl utilities is provided in the Parallels Server
Bare Metal 5.0 Command Line Reference Guide and their manual pages.
Managing Adapters in Containers
Parallels Server Bare Metal provides you with ample opportunities of configuring veth virtual
network adapters in Containers and including them in different network environments. This section
shows you the way to perform the following operations:
•
Create new virtual network adapters in Containers and delete existing ones.
•
Configure the parameters of an existing virtual network adapter.
•
Join Container virtual network adapters to virtual networks.
All these operations are described in the following subsections in detail.
179
Managing Parallels Server Bare Metal Network
Creating and Deleting veth Network Adapters
By default, any Container on the Parallels server starts functioning in the venet0 mode right after
its creation. However, at any time you can create additional virtual adapters for Containers and set
them to work in the virtual network mode. You can do this using the --netif_add option of the
pctl set command.
Let us assume that you wish to create a new virtual adapter with the name of eth1 in Container
101 and make it function in the virtual network mode. To do this, run the following command :
# pctl set 101 --netif_add eth1 --save
Saved parameters for Container 101
The settings of the newly created virtual adapter are saved as the value of the NETIF parameter in
the configuration file of Container 101 (/etc/vz/conf/101.conf). So, you can use the following
command to display the parameters assigned to the veth network adapter in Container 101:
# grep NETIF /etc/vz/conf/101.conf
NETIF="ifname=eth1,mac=00:10:41:F0:AA:B6,host_mac=00:18:51:A0:8A:D7"
As you can see, the parameters set for the veth virtual network adapter during its creation are the
following:
•
ifname: the name set for the veth Ethernet interface in Container 101. You specified this
name when creating the Container virtual network adapter. Usually, names of Ethernet
interfaces in Containers are set in the form of ethAd_N where Ad_N denotes the index number
of the created adapter (for example, eth0 or eth1). However, you can choose any other name
you like and specify it during the virtual adapter creation.
•
mac: the MAC address assigned to the veth Ethernet interface in Container 101.
•
host_mac: the MAC address assigned to the veth Ethernet interface on the Parallels server.
ifname is the only mandatory parameter that you need to specify when creating a Container virtual
network adapter. All the other parameters are optional and generated by Parallels Server Bare
Metal automatically, if not indicated.
At any time, you can remove the veth virtual network adapter from Container 101 by executing the
following command:
# pctl set 101 --netif_del eth1 --save
Saved parameters for Container 101
180
Managing Parallels Server Bare Metal Network
Configuring veth Adapter Parameters
While functioning in the virtual network mode, each Container virtual network adapter appears as a
full participant on the network to which it is connected and needs to have its own identity on this
network.
Fist of all, to start functioning on a TCP/IP network, a veth virtual adapter should be assigned an
IP address. This can be done as follows:
# pctl set 101 --ifname eth1 --ipadd 192.168.144.123 --save
Saved parameters for Container 101
This command sets an IP address of 192.168.144.123 for the eth1 adapter in Container 101.
If you want to use the Dynamic Host Configuration Protocol (DHCP) to make the eth1 adapter of
Container 101 automatically receive TCP/IP configuration settings, you can issue the following
command instead:
# pctl set 101 --ifname eth1 --dhcp yes --save
Saved parameters for Container 101
Any static IP address assigned to the Container virtual network adapter can be removed by
executing the following command:
# pctl set 101 --ifname eth1 --ipdel 192.168.144.123 --save
Saved parameters for Container 101
You can also delete all IP addresses set for Container 101 at once:
# pctl set 101 --ifname eth1 --ipdel all --save
Saved parameters for Container 101
You may also wish to set the following parameters for a Container network adapter:
•
A DNS server that the Container virtual adapter is supposed to use:
# pctl set 101 --ifname eth1 --nameserver 192.168.100.111 --save
Saved parameters for Container 101
•
A gateway to be used for routing the traffic of the Container virtual adapter:
# pctl set 101 --ifname eth1 --gateway 192.168.111.1 --save
Saved parameters for Container 101
Detailed information on all options which can be used with the pctl set command to manage
Container adapter parameters is given in the Parallels Server Bare Metal 5.0 Command Line
Reference Guide and the pctl manual pages.
181
Managing Parallels Server Bare Metal Network
Connecting Containers to Virtual Networks
With the implementation of veth virtual adapters allowing Containers to function as full participants
on the network, it has become possible to include Containers in a wide range of network
configurations the most common of which are Ethernet networks and VLANs (virtual local area
networks). The process of connecting veth virtual network adapters to an Ethernet network or to a
VLAN is carried out using certain physical and VLAN adapters, respectively, available on the server
and involves completing the following tasks:
1
Creating a virtual network that will act as an intermediary between the veth adapters and the
physical/VLAN adapter.
2
Connecting the veth virtual adapter you want to include in an Ethernet network/VLAN to the
virtual network.
3
Joining the virtual network where the veth virtual adapters are included to the corresponding
physical/VLAN adapter.
After completing these tasks, the Container virtual network adapters will be able to communicate
with any computer on the network (either Ethernet or VLAN) where they are included and have no
direct access to the computers joined to other networks.
The process of creating new virtual networks and joining physical and VLAN adapters to them is
described in the Creating a Virtual Network (p. 175) and Connecting Virtual Networks to
Adapters (p. 178) sections, respectively. In the example below we assume the following:
•
The eth0 physical adapter and the psbmnet1 virtual network exist on the server.
•
The eth0 physical adapter is connected to the local Ethernet network and to the psbmnet1
virtual network.
•
You want to connect Container 101 and Container 102 to the local Ethernet network.
To join Container 101 and 102 to the local Ethernet network behind the eth0 adapter, you need
connect these Containers to the psbmnet1 virtual network. To do this:
1
Find out the name of the veth Ethernet interfaces in Container 101 and 102:
# vzlist -a -o ctid,ifname
CTID IFNAME
101 eth1
102 eth0
103 -
The command output shows that the veth Ethernet interfaces in Container 101 and 102 have
the names of eth1 and eth0, respectively.
Note: To add a veth adapter to a virtual network, you must use the name of its Ethernet interface in
the Container.
2
Join the veth adapters to the psbmnet1 virtual network:
• Add the veth adapter of Container 101 to the virtual network:
# pctl set 101 --ifname eth1 --network psbmnet1 --save
182
Managing Parallels Server Bare Metal Network
Saved parameters for Container 101
• Add the veth adapter of Container 102 to the virtual network:
# pctl set 102 --ifname eth0 --network psbmnet1 --save
Saved parameters for Container 102
After completing these tasks, Container 101 and Container 102 will be able to access any of the
servers in the network where the eth0 physical adapter is connected.
At any time, you can disconnect the veth virtual network adapters of Container 101 and 102 from
the psbmnet1 virtual network by executing the following commands:
•
To disconnect the veth adapter of Container 101 from the virtual network:
# pctl set 101 --ifname eth1 --network "" --save
Saved parameters for Container 101
•
To disconnect the veth adapter of Container 102 from the virtual network:
# pctl set 102 --ifname eth1 --network "" --save
Saved parameters for Container 102
Managing Adapters in Virtual Machines
This section provides information on how you can manage virtual network adapters in virtual
machines. You will learn to do the following:
•
Create new virtual network adapters and delete existing ones.
•
Configure the parameters of an existing virtual network adapter.
•
Join virtual network adapters to virtual networks.
All these operations are described in the following subsections in detail.
183
Managing Parallels Server Bare Metal Network
Creating and Deleting Virtual Adapters
A virtual machine can have up to 16 virtual network adapters. Each adapter can be connected to a
different network. Let us assume that you want to create a new virtual adapter for the MyVM virtual
machine. To do this, you can execute the following command :
# pctl set MyVM --device-add net
Creating net1 (+) type=host-only iface='default' mac=001C42AF3D69
The VM has been successfully configured.
To check that the network adapter (net1) has been successfully added to the virtual machine, run
this command:
# pctl list --info MyVM
ID: {f3b3d134-f512-324b-b0b1-dbd642f5220b}
Name: Windows XP
...
net0 (+) type=host-only iface='default' mac=001C42566BCF
net1 (+) type=host-only iface='default' mac=001C42AF3D69
At any time, you can remove the newly created network adapter (net1) by executing the following
command:
# pctl set MyVM --device-del net1
Remove the net1 device.
The VM has been successfully configured.
For the full of options that can be used when creating a new virtual network adapter, see the
Parallels Server Bare Metal 5.0 Command Line Reference Guide.
184
Managing Parallels Server Bare Metal Network
Configuring Virtual Adapter Parameters
Parallels Server Bare Metal allows you to configure the following parameters of virtual machine
adapters:
Configuring MAC Addresses
If you need for some reason to regenerate the current MAC address of a network adapter, you can
use the following command:
# pctl set MyVM --device-set net1 --mac 00:1C:42:2D:74:00
Creating net1 (+) network=Bridged mac=001C422D7400
The VM has been successfully configured.
This command sets the MAC address of 00:1C:42:2D:74:00 for the net1 adapter in the MyVM
virtual machine. If do not know what MAC address to assign to your virtual adapter, you can make
pctl set automatically generate a new MAC address. To do this, run the following command:
# pctl set MyVM --device-set net1 --mac auto
Creating net1 (+) network=Bridged mac=001C42C84F3E
The VM has been successfully configured.
Configuring IP Parameters
As any other standalone server, each virtual machine must have a number of TCP/IP settings
configured in the proper way to successfully operate on the network. These settings include:
•
IP address
•
default gateway
•
DNS server
Usually, you define all these settings when you create the virtual machine. However, if you have not
yet set any of the settings or want to modify any of them, you can use the pctl set command.
For example, you can execute the following command to assign the IP address of
192.129.129.20 to the net1 adapter in the MyVM virtual machine, set the default gateway to
192.129.129.1 and the DNS server to 192.192.192.10:
# pctl set MyVM --device-set net1 --ipadd 192.129.129.20 --gw 192.129.129.1 -nameserver 192.192.192.10
Along with a static assignment of network parameters to a virtual adapter, you can make the
adapter receive its TCP/IP settings automatically using the Dynamic Host Configuration Protocol
(DHCP). For example, you can run this command to make the net1 adapter in the MyVM virtual
machine get its IP settings through DHCP:
# pctl set MyVM --device-set net1 --dhcp yes
Creating net1 (+) network=Bridged mac=001C42C84F3E
Enable automatic reconfiguration for this network adapter.
The VM has been successfully configured.
185
Managing Parallels Server Bare Metal Network
Notes:
1. You can configure the network parameters only of those virtual machines that have Parallels Tools
installed.
2. Detailed information on all options which can be used with the pctl set command to manage virtual
machine adapter parameters is given in the Parallels Server Bare Metal 5.0 Command Line Reference
Guide and the pctl manual pages.
Connecting Virtual Machines to Virtual Networks
In Parallels Server Bare Metal, you can connect virtual machines to virtual networks of the following
types:
•
Bridged. This type of virtual network allows the virtual machine to use one of the physical
server's network adapters, which makes it appear as a separate computer on the network the
corresponding adapter belongs to.
•
Host-only. This type of virtual network allows the virtual machine to access only the Parallels
server and the virtual machines joined to this network.
By default, any newly created adapter is connected to the Bridged network. To join a virtual
machine to another network, use the pctl set command. For example, the following session
demonstrates how you can connect the net0 adapter of the MyVM virtual machine to the
psbmnet1 virtual network.
Before connecting the MyVM virtual machine to the psbmnet1 virtual network, you may wish to
check the network adapter associated with this virtual network. You can do it, for example, using
the following command:
# prlsrvctl net list
Network ID
Type
Host-Only
host-only
Bridged
bridged
psbmnet1
bridged
Bound To
eth0
eth1
From the command output, you can see that the psbmnet1 virtual network is attached to the
eth1 physical adapter on the Parallels server. That means that, after connecting the MyVM virtual
machine to the psbmnet1 virtual network, the virtual machine will be able to access all computers
on the network where the eth1 adapter is connected.
Now you can run the following command to join the net1 adapter of the MyVM virtual machine to
the psbmnet1 virtual network:
# pctl set MyVM --device-set net0 --network psbmnet1
Creating net0 (+) network=psbmnet1 mac=001C422D7493
The VM has been successfully configured.
186
Managing Parallels Server Bare Metal Network
Managing Private Networks
This section describes how to manage private networks and subnetworks in Parallels Server Bare
Metal 5.0.
Learning Private Networks
By default, all virtual machines and Containers on the physical server operating in the host-routed
mode can access each other even if you connect them to different subnets. For example, if
Container 101 has the IP address of 100.10.10.101 and the MyVM virtual machine has the IP
address of 100.10.11.102 and you set the subnet mask for them to 255.255.255.0, the virtual
machine and Container will be able to communicate with each other, though they technically
belong to different subnets: 100.10.10.0 and 100.10.11.0.
Note: You can also include virtual machines and Containers operating in the virtual network mode in
private networks. For information on how you can do this, see Setting Up Private Networks (p. 191).
In Parallels Server Bare Metal 5.0, you can create the so-called private networks. Within these
private networks, you can make subnets and connect virtual machines and Containers to these
subnets so that the virtual machines and Containers from one subnet will not be able to access
virtual machines and Containers from the other subnets, virtual machines and Containers outside
the private network, and computers on external networks. The following figure demonstrates a
system containing a private network:
187
Managing Parallels Server Bare Metal Network
In this example, the network is configured as follows:
•
A private network (Private Network) is created within the physical server network (Host
Network).
•
The private network contains two private subnets: Subnet 1 and Subnet 2.
•
Container 101 and VM1 are connected to Subnet 1, and Container 103 and VM2 are joined to
Subnet 2.
•
Container 105 and VM3 do not belong to the private network.
•
The physical server network is connected to an external network (External Network) that
contains computers Computer 1, Computer 2, and Computer 3.
In this network, Container 101 and VM1 can access each other, but cannot connect to Container
103, Container 105, VM2, and VM3. Container 103 and VM2, in turn, can also access each other,
but cannot connect to Container 101, Container 105, VM1, and VM3. None of the virtual machines
and Containers in the private network can access computers on the external network.
Network Across Several Nodes
The example above deals with a private network created within one physical server. However,
private networks can span virtual machines and Containers on two or more servers. The following
figure demonstrates such a network:
188
Managing Parallels Server Bare Metal Network
In this figure, the private network also includes two private subnets—Subnet 1 and Subnet 2, but
the virtual machines and Containers included in these subnets reside on two physical servers.
Container 101 and VM1 are joined to Subnet 1, and Container 102, Container 203, and VM2 are
joined to Subnet 2. The virtual machine and Container on Subnet 1 can connect to each other but
cannot access the virtual machines and Containers on Subnet 2, and vice versa.
Weak Private Networks
By default, when you create a private network, no virtual machine or Container on this network can
access
•
virtual machines and Containers that are joined to other subnets in the private network
•
virtual machines and Containers that are not part of the private network
•
computers that are located on external networks
However, you can configure a private network so that its virtual machines and Containers cannot
communicate with virtual machines and Containers on other subnets in the private network, but
can connect to virtual machines and Containers outside the private network and to computers on
external networks. Such private networks are called weak private networks. "Weak" in this context
means that these networks can be accessed by computers on external networks and are,
therefore, more prone to security vulnerabilities and threats. The following picture demonstrates a
system with a weak private network:
189
Managing Parallels Server Bare Metal Network
In this example, the private network on the physical server is divided into two subnets: Subnet 1
and Subnet 2. Container 101 and VM1 are connected to Subnet 1, and Container 103 and VM2
are joined to Subnet 2. Container 105 and VM3 do not belong to the private network. Container
101 and VM1 can access each other, but cannot connect to Container 103 and VM2. Container
103 and VM2, in turn, can also access each other, but cannot connect to Container 101 and VM1.
All virtual machines and Containers in the private network can communicate with Container 105
and VM3 and, as they have public IP addresses assigned, can also access computers on other
networks (for example, the computers Computer 1 and Computer 2 on the external network
External Network). To protect the virtual machines and Containers from possible security
vulnerabilities and threats, the firewall is configured on the physical server, blocking unauthorized
access to the virtual machines and Containers.
190
Managing Parallels Server Bare Metal Network
Setting Up Private Networks
Before you start setting up a private network, you need to decide on the following:
1
The name to assign to the private network.
2
The range of IP addresses to allocate to the private network.
3
The number of subnets and hosts in the private network.
Let us assume that you want to create two private networks with the following parameters:
•
The first network has the name privnet1 and includes the IP addresses from 10.10.0.0
through 10.10.255.255.
•
The second network has the name privnet2 and includes the IP addresses from 10.11.0.0
through 10.11.255.255.
To create these two networks, you can run the following commands on the physical server:
# prlsrvctl privnet add privnet1 --ipadd 10.10.0.0/16
# prlsrvctl privnet add privnet2 --ipadd 10.11.0.0/16
Now if you assign to one virtual machine or Container an IP address from the range 10.10.0.0
through 10.10.255.255 and to another virtual machine or Container from the range 10.11.0.0
through 10.11.255.255, they will not be able to access each other because they will belong to
different private networks.
If you want to create a private network for several virtual machines and Containers only, you can
omit the network mask and specify only the required IP addresses:
# prlsrvctl privnet add privnet3 --ipadd 10.12.0.101
# prlsrvctl privnet add privnet3 --ipadd 10.12.0.102
The commands above create the private network privnet3 and include only two IP addresses in
this network: 10.12.0.101 and 10.12.0.102. You can then assign these IP address to the two of
your virtual machines and Containers, thus isolating their network traffic from that of other virtual
machines and Containers and external computers.
Enabling Private Networks for virtual machines and Containers Operating in Virtual
Network Mode
By default, you can include in private networks only virtual machines and Containers operating in
the host-routed mode. If you want to connect to a private network some of virtual machines and
Containers operating in the virtual network mode, you need first to enable the private network
support on the physical server. To do this, change the value in the
/proc/sys/net/vzpriv_handle_bridge file from 0 to 1:
# echo 1 > /proc/sys/net/vzpriv_handle_bridge
Note: Enabling the support for private networks may affect the network performance of virtual machines
and Containers that operate in the virtual network mode and are assigned IPv4 addresses.
191
Managing Parallels Server Bare Metal Network
Creating Weak Private Networks
In a weak private network, any virtual machine or Container on the network can communicate with
the other virtual machines and Containers in the same subnet, virtual machines and Containers
outside the private network, and computers on external networks. To create a weak private
network, you can run the following command:
# prlsrvctl privnet set privnet_name --ipadd '*'
where privnet_name is the name of the private network you want to configure as a weak one.
For example, to configure the privnet1 private network as a weak one, execute the following
command:
# prlsrvctl privnet set privnet1 --ipadd '*'
To make sure that privnet1 is now a weak network, check the contents of the
/proc/vz/privnet/sparse file:
# cat /proc/vz/privnet/sparse
1: * 10.10.0.0/16
2:
10.11.0.0/16
The asterisk before 10.10.0.0/16 denotes that privnet1 is now acting as a weak network. To
revert the changes made to the privnet1 file, run this command:
# prlsrvctl privnet set privnet1 --ipdel '*'
Connecting virtual machines and Containers to Private Subnets
Once you set up a private network, you can connect Containers to different subnets within this
network. Assuming that you followed the instructions above, you now have two private networks.
The privnet1 network includes the IP addresses from 10.10.0.0 through 10.10.255.255, and the
privnet2 network contains the IP addresses from 10.11.0.0 through 10.11.255.255. Let us join
Container 101 to privnet1 and the MyVM virtual machine to privnet2. To do this:
1
Assign IP address 10.10.10.101 to Container 101:
# pctl set 101 --ipadd 10.10.10.101 --save
2
Assign the IP address of 10.11.10.101 to the MyVM virtual machine:
# pctl set MyVM --device-set net0 --ipadd 10.11.10.101
Now Container 101 and the MyVM virtual machine belong to different subnets and cannot access
each other.
Removing Private Networks
At any time, you can remove privnet1 and privnet2 by running these commands:
# prlsrvctl privnet del privnet1
# prlsrvctl privnet del privnet2
Once you execute these commands, all virtual machines and Containers that were included in the
privnet1 network should be able to connect to the virtual machines and Containers that were
joined to the privnet1 network.
192
CHAPTER 7
Managing Licenses
The given chapter provides information on managing Parallels Server Bare Metal licenses. In
particular, you will know how to view the current license status, to install a new license on your
server or to update an existing one, to transfer the license from one server to another, etc.
In This Chapter
Installing the License .............................................................................................. 194
Updating the Current License ................................................................................. 195
Transferring the License to Another Server .............................................................. 196
Viewing the Current License ................................................................................... 197
Managing Licenses
Installing the License
Depending on the way you have obtained your Parallels Server Bare Metal license, the process of
installing the license slightly differs:
•
If you have obtained the license in the form of a product key, you can install it on the server
using the -p option of the vzlicload command. For example, you can execute the following
command to install the XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX product key:
# vzlicload -p 5BVMF2-560MM0-D28DQA-B59NTE-10H4HG
Processing product key "XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX"...
License VZSRV was loaded successfully
--1 of 1 licenses was loaded
Note: You can also use the vzlicload utility to upgrade the license. For example, this may be
necessary if your current license does not support using Parallels Virtual Automation for managing
Parallels servers.
•
If you have obtained the license in the form of an activation code, you can install it on the server
using the -a option of the vzlicupdate command. For example:
# vzlicupdate -a XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX
where XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX is your activation code. When executed,
vzlicupdate connects to the Parallels Key Authentication (KA) licensing server and transmits
the specified activation code there. The licensing server, in turn, generates a license file, sends it
back to the server from where the activation code has been dispatched, and automatically
installs it on this server. So, before executing the aforementioned command, make sure that
your Parallels server is connected to the Internet.
If you are activating your installation by means of an activation key, you must have an active Internet
connection to successfully complete the license installation. Otherwise, you will be presented with
the corresponding warning message informing you of the steps you have to take to activate your
license. As a rule, these steps are the following:
1
Visiting the http://www.parallels.com/en/support/virtuozzo/activate web page and activating the
license manually.
2
Providing the following information on this web page:
• In the Product Code field, specify your license activation code.
• In the HWID field, provide the ID of your server.
• In the Enter following digits field, type the digits displayed next to this field.
3
Clicking the ACTIVATE LICENSE button.
If you have entered the correct information on the Virtuozzo License Activation page, you will be
provided with a link to a license file that you should download to and install on the server. For
example, you can run this command to install the obtained license file
# vzlicload -f /etc/vzlicense
194
Managing Licenses
This command will install the license file with the name of vzlicense on your server.
Updating the Current License
In Parallels Server Bare Metal, you can use the vzlicupdate utility to update the license currently
installed on the Parallels server. When executed, the utility tries to connect to the Parallels Key
Authentication (KA) server and to retrieve a new license and install it on the server. To update your
license, do the following:
1
Make sure that the Parallels server where you wish to update the license is connected to the
Internet.
2
Execute the following command on the server:
# vzlicupdate
Start updating license [6E62.3D01.6BEC.E8D7.CE42.4517.68CB.E102]
...
By default, vzlicupdate tries to access the KA server having the hostname of
ka.parallels.com. However, you can explicitly specify what KA server to use using the -server option:
# vzlicupdate --server ka.server.com
In this case, the vzlicupdate utility will try to connect to the KA server with the hostname of
ka.server.com, to get a new license from this server, and to install it on the server where
vzlicupdate has been executed.
Note: Your physical server must be assigned at least one public IPv4 address for the correct operation of
the vzlicupdate utility.
195
Managing Licenses
Transferring the License to Another Server
Sometimes, you may wish to transfer licenses from one Parallels server (source server) to another
(destination server). For example, this may be the case if the server where the license is installed
starts experiencing problems or requires the hardware upgrade.
The procedure of transferring a license from one Parallels server to another depends on the license
type and can be one of the following:
Activation with a product key
If you have activated your Parallels Server Bare Metal installation by means of a product key, you
can transfer the installed license from the source to the destination server as follows:
1
Remove the installed license from the source server (e.g., using the vzlicload -r
product_key command).
2
Log in to the destination server.
3
Install the product key on the destination server. Detailed information on how to install Parallels
Server Bare Metal licenses is provided in Installing a License (p. 194).
Activation with an activation code
If you have activated your Parallels Server Bare Metal installation by means of an activation code,
you can use the vzlicupdate utility to move licenses between Parallels servers. For example, to
transfer a license that has been installed using the XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX
activation code, do the following:
1
Ascertain that the source server is shut down, or the license is removed from this server.
2
Make sure that the destination server is up and connected to the Internet.
3
Log in to the destination server (e.g., via ssh).
4
Execute the following command on the destination server:
# vzlicupdate -t -a XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX
When executed, vzlicupdate sends the activation code to the Parallels KA server, thus
informing the server of its intention to transfer the license to a new Parallels server. The KA server
verifies the received code, generates a new license file, sends it back to the destination server, and
installs it there.
Note: Your physical server must be assigned at least one public IPv4 address for the correct operation of
the vzlicupdate utility.
You can check that the license transferal has completed successfully using the vzlicview utility.
For example:
# vzlicview
Show installed licenses...
196
Managing Licenses
VZSRV
status="ACTIVE"
version=X.X
serial="XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX"
expiration="05/01/2012 23:59:59"
...
Detailed information on the vzlicview utility and its output is provided in Viewing Current
License (p. 197).
Viewing the Current License
The given subsection familiarizes you with the way to view the information on the license installed
on your Parallels server.
197
Managing Licenses
Viewing the License
In Parallels Server Bare Metal, you can use the vzlicview utility to view the information on the
installed license and find out its current status. When executed, this utility processes the license
currently installed on the Parallels server and prints the license contents along with its status. A
sample output of vzlicview is given below:
# vzlicview
Searching for installed licenses...
VZSRV
status="ACTIVE"
version=X.X
serial="XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX"
expiration="12/01/2012 23:59:59"
graceperiod=86400 (86400)
key_number="PSBM.00000001.0000"
cpu_total=64 (1)
ct_total=100 (1)
max_vzmcpmc_users=128
max_pim_users=260
platform="Any"
product="PSBM"
vzpp_allowed=1
backup_mgmt_allowed=1
workflow_mgmt_allowed=1
vzagent_allowed=1
nr_vms=10 (2)
architecture="Any"
The command output shows the full information about the license. The main license parameters are
listed in the following table:
Column Name
Description
status
The license status. The information on all possible license statuses
is provided in License Statuses (p. 200).
version
The version of Parallels Server Bare Metal for which the license was
issued.
serial
The license serial number.
expiration
The license expiration date, if it is time-limited.
The period, in seconds, during which Parallels Server Bare Metal
continues functioning if
graceperiod
•
the license has expired
•
the number of running virtual machines and Containers
exceeds the limit defined by the license
key_number
The number under which the license is registered on the Parallels
Key Authentication server.
cpu_total
The total number of CPUs you are allowed to install on the Parallels
server.
ct_total
The total number of Containers you are allowed to simultaneously
run on the Parallels server.
198
Managing Licenses
max_vzmc_users
The number of users able to simultaneously connect to the server
using Parallels Management Console.
max_vzcc_users
The number of users able to simultaneously connect to the server
using Parallels Virtual Automation (formerly, Parallels Infrastructure
Manager).
platform
The operating system with which the license is compatible.
product
The product name for which the license has been issued.
Indicates whether you can manage virtual machines and Containers
using Parallels Power Panel:
vzpp_allowed
•
1: the 'Parallels Power Panel' functionality is enabled
•
0: the 'Parallels Power Panel' functionality is disabled
Indicates whether the 'backup' functionality is enabled for the given
server:
backup_mgmt_allowed
•
1: the 'backup' functionality is enabled
•
0: the 'backup' functionality is disabled
Indicates whether the 'virtual machine and Container requesting'
functionality is enabled for the given server:
workflow_mgmt_allowed
•
1: the 'virtual machine and Container requesting'
functionality is enabled
•
0: the 'virtual machine and Container requesting'
functionality is disabled
Indicates whether you are allowed to use the Parallels Agent
functionality on the given server:
vzagent_allowed
•
1: the Parallels Agent functionality is enabled
•
0: the Parallels Agent functionality is disabled
nr_vms
The number of virtual machines you are allowed to simultaneously
run on the Parallels server.
architecture
The system architecture with which the license is compatible.
concerto
If this field is present, the license supports the ability to use the
Plesk application in Containers.
Licenses with a combined limit on virtual machines and Containers
Some licenses shipped with Parallels Server Bare Metal 5.0 define a combined limit on the number
of virtual machines and Containers you are allowed to simultaneously run on the Parallels server
rather than set limits separately for virtual machines and Containers. In this case, the license output
is as follows:
# vzlicview
Searching for installed licenses...
VZSRV
status="ACTIVE"
version=X.X
serial="XXXXXX-XXXXXX-XXXXXX-XXXXXX-XXXXXX"
expiration="12/01/2012 23:59:59"
graceperiod=86400 (86400)
199
Managing Licenses
key_number="PSBM.00000001.0000"
cpu_total=64 (1)
ct_total=100 (1)
max_vzmcpmc_users=128
max_vzcc_users=260
platform="Any"
product="PSBM"
vzpp_allowed=1
backup_mgmt_allowed=1
workflow_mgmt_allowed=1
vzagent_allowed=1
nr_vms="combined" (2)
servers_total=100
architecture="Any"
As you can see, the output now contains one more parameter—servers_total. This parameter
defines the total number of virtual machines and Containers you can simultaneously run on the
Parallels server. For example, according to the license above, you can run 100 Containers, or 100
virtual machines, or 50 Containers and 50 virtual machines on the server at the same time.
License Statuses
When viewing information on your license, pay special attention to the license status that can be
one of the following:
ACTIVE
The license installed on the server is valid and active.
VALID
The license the utility parses is valid and can be installed on the server.
EXPIRED
The license has expired and, therefore, could not be installed on the server.
GRACED
The license has been successfully installed on the server; however, it is currently on the
grace period because
INVALID
200
•
the license has expired
•
the number of running virtual machines and Containers exceeds the limit
defined by the license
The license is invalid (for example, because of the server architecture mismatch) or
corrupted.
CHAPTER 8
Keeping Your System Up To Date
This chapter explains the ways to keep your Parallels server up to date. The components you need
to take care of are the following:
•
Parallels Server Bare Metal software
•
virtual machines and Containers created on the Parallels server
In This Chapter
Updating Parallels Server Bare Metal Software ........................................................ 202
Updating Software In Virtual Machines..................................................................... 210
Updating Containers ............................................................................................... 210
Keeping Your System Up To Date
Updating Parallels Server Bare Metal Software
Parallels Server Bare Metal is constantly developing: there appear new versions of the Parallels
Server Bare Metal core and of existing utilities, OS and application templates are perfected, new
templates and utilities are also added from time to time. Thus, Parallels Server Bare Metal may
sometimes be repackaged to include the latest changes in any of its parts. As these changes grow
in number, new product versions are released.
Parallels Server Bare Metal provides a special utility, vzup2date, allowing you to easily and quickly
update your Parallels server. The main components that need to be updated are the following:
•
Parallels Server Bare Metal system software (packages built by Parallels)
•
Parallels Server Bare Metal templates
The vzup2date utility can be launched in two modes:
•
Graphical mode. In this mode, you use a special wizard to update either the Parallels Server
Bare Metal system files or templates depending on the options passed to vzup2date.
•
Command line mode containing two submodes:
• the batch submode
• the messages submode
In comparison to the graphical mode, the command line mode provides more possibilities for
the Parallels Server Bare Metal updates management (e.g. the ability to use special filters while
selecting updates for your system).
Both modes are described in the following subsections in detail.
Note: Your physical server must be assigned at least one public IPv4 address for the correct operation of
the vzup2date utility.
202
Keeping Your System Up To Date
Updating in Graphical Mode
In the graphical mode, the vzup2date utility can be launched in two submodes. If invoked without
any parameters or with the -s switch, it is supposed to check and, if necessary, download and
install Parallels Server Bare Metal system files. On the other hand, specifying the -z option when
invoking the utility tells it to perform the same operations for OS and application EZ templates.
There is no single interface for checking system files and templates at once, as these operations
are different in nature. Therefore, you should consecutively call the vzup2date utility with and
without the -z option, if you wish to check for all available system and template updates.
Note: You can explicitly specify that the vzup2date utility is to be run in the graphical mode by passing
the -m interactive switch to it.
The vzup2date utility is implemented as a wizard, the first few steps of which are common for
both modes. After you launch the utility from the command line, you will be presented with the
Welcome screen. In this window, you can do one of the following:
•
Click the Next button to connect to the Parallels default repository.
•
Click the Configure button to display the current settings used to connect to the repository
housing Parallels Server Bare Metal updated packages and templates and to configure it, if
necessary:
Figure 1: Updating System - Specifying Repository
203
Keeping Your System Up To Date
The information on this screen is taken from the
/etc/sysconfig/vzup2date/vzup2date.conf file on the Parallels server. If you wish to
change this information and save the changes to the configuration file, enter the correct settings
into the fields provided, and press OK.
Once you press Next in the Welcome window, the utility will try to connect to the specified
repository (either the Parallels default repository or your own one) and, if the connection is
successful, display the next screen, which will vary depending on the mode of the vzup2date
invocation. First, we will describe the mode of updating Parallels Server Bare Metal system files and
then proceed with updating your EZ templates.
Note: The vzup2date utility might see that the selected update includes an updated version of the
vzup2date utility itself. In this case you will first have to perform an update of this utility and then to relaunch it and select the desired Parallels Server Bare Metal system update once again.
Updating System Files
After the vzup2date utility has checked the repository and found any updates, you are presented
the following window:
204
Keeping Your System Up To Date
This window displays the list of updates that can be installed on your Parallels server. If you want to
update to the latest Parallels Server Bare Metal core and utilities versions, just press Next on this
screen, and the vzup2date utility will download and install them asking your confirmation before
each action.
On the other hand, if you have a reason not to install the latest updates for both the Parallels Server
Bare Metal core and utilities, press Customize. In this case, you will be able to choose whether to
perform customization on the Parallels Server Bare Metal core or utilities. This step will be skipped if
updates are currently available either only for the core or only for utilities. On the next step, you will
be asked to choose the desired core or utilities updates, in case there are many.
Notice that the bottommost update includes the functionality of all the other updates. You can
select any of the intermediary updates and press Select to go back to the List of Selected
Updates screen and read the information on this update. You will be able to perform customization
more than once until you finally decide on the set of updates to be applied and press Next.
Downloading and installing the necessary updates is straightforward.
205
Keeping Your System Up To Date
Updating EZ Templates
Updating EZ templates consists in updating one or more EZ templates configuration files located in
the /vz/template/<os_name>/<os_version>/<arch>/config directory on the Parallels
server and takes place if you have launched the vzup2date utility with the -z option. The first few
steps of the wizard were described in the Updating in Graphical Mode subsection (p. 203). As
soon as you press Next in the Welcome... window, the utility will try to connect to the EZ
templates repository (either the Parallels default repository or your own one) and, if the connection
is successful, display the EZ Templates Selection window listing all EZ templates that have one or
more updates available or that are not installed on your server at all.
In this window, you can do one of the following:
•
If you wish to download and install all available EZ templates/template updates for a certain
Linux distribution, select this distribution by placing the cursor beside it and pressing the space
bar on your keyboard; then click Next.
206
Keeping Your System Up To Date
•
If you wish only certain EZ templates of the corresponding Linux distribution to be
installed/updated on the Parallels server, place the cursor beside this distribution and press F2
on your keyboard. You will be presented with the Templates selection window where you can
select the corresponding EZ templates.
After choosing the right EZ templates, click the Select button to close the displayed window,
and then click Next to proceed with the wizard.
Note: New application EZ templates for a Linux distribution can be installed on the Parallels server only
if the corresponding OS EZ template is already installed on this server.
In the next step, you can review the EZ templates/template updates you selected on the previous
step and scheduled for downloading and installing on your server. If you are not satisfied with the
chosen templates/template updates, click the Back button to return to the previous step and
modify the set of templates; otherwise, click Next to start downloading the templates/template
updates.
207
Keeping Your System Up To Date
After the EZ templates/templates have been successfully downloaded to the server, the Installing
EZ template window is displayed.
In this window, you can view the templates/template updates ready to be installed on your server. If
you are installing a new OS EZ template/OS EZ template update, you can select the Run vzpkg
cache after installation option and specify whether to cache the corresponding OS EZ
template/template update right after its installation on the server or to do it at a later time. By
default, all OS EZ templates are just installed on the Parallels without being cached. However, you
can select the provided check box and schedule your OS EZ template/template update for
caching. Clicking Next starts installing the EZ templates on the server. By the time the wizard
finishes, you should have updated OS and application templates on your system.
208
Keeping Your System Up To Date
Updating in Command-Line Mode
Another way of updating your Parallels Server Bare Metal system files and templates is to run the
vzup2date utility in the command line mode and to pass the corresponding commands,
switches, and options to it. While executing vzup2date in the command line mode, you can
choose between the batch and messages submodes. Both submodes can be used to update
either the Parallels Server Bare Metal system files or EZ templates and have the identical syntax.
However, the output produced by these commands is different. The messages submode output is
less user friendly than that of the batch submode and is mostly suitable for machine processing.
To run the vzup2date utility in the command line mode, you can use either the -m batch switch
or the -m messages switch intended for executing vzup2date in the batch and messages
submodes, respectively.
Let us assume that you wish to update Parallels Server Bare Metal system files by installing the
latest core in the batch submode. To do this, you can issue the following command on the Parallels
server:
# vzup2date -m batch install --core
This will check the Parallels Server Bare Metal repository for the latest core updates and, in the
case of finding any, download and install them on your server.
To update your Parallels Server Bare Metal installation, you may need to edit the
/etc/sysconfig/vzup2date/vzup2date.conf file to specify the repository from where the
updates are to be downloaded or configure a number of other parameters. Detailed information on
the vzup2date.conf file is provided in the Parallels Command Line Reference Guide.
You can also execute the vzup2date utility in the batch mode to update the EZ templates
installed on the Parallels server. For example, you can issue the following command
# vzup2date -t -m batch install --all-os
to update all OS templates installed on your server. Detailed information on all options that can be
passed to the vzup2date utility is given in the Parallels Command Line Reference Guide.
Note: To perform the aforementioned operations in the messages submode, you should pass the -m
messages option to the vzup2date utility instead of -m batch.
209
Keeping Your System Up To Date
Updating Software In Virtual Machines
To keep software in your virtual machines up to date, you can use the same means you would use
on standalone computers running the corresponding operating systems:
•
In Linux-based virtual machines, you can use the native Linux updaters (up2date, yum, or
yast).
•
In Windows-based virtual machines, you can use the native Windows updaters (e.g. the
Windows Update tool).
You should regularly run these updaters to ensure that your system has the latest updates and
fixes (including security patches) installed. For more information on native updaters, refer to the
documentation shipped with your operating system.
Updating Containers
Parallels Server Bare Metal provides two facilities to keep your Containers up to date. These
facilities include:
•
Updating EZ templates software packages inside a particular Container by means of the vzpkg
utility. Using this facility, you can keep any of the Containers existing on your Parallels server up
to date.
•
Updating caches of the OS EZ templates installed on the Parallels server. This facility allows you
to create new Containers already having the latest software packages installed.
210
Keeping Your System Up To Date
Updating EZ Template Packages Inside a Container
Parallels Server Bare Metal allows you to update packages of the OS EZ template a Container is
based on and of any application EZ templates applied to the Container. You can do it by using the
vzpkg update utility. Assuming that Container 101 is based on the redhat-el5-x86 OS EZ
template, you can issue the following command to update all packages included in this template:
# vzpkg update 101 redhat-el5-x86
...
Updating: httpd
####################### [1/4]
Updating: vzdev
####################### [2/4]
Cleanup : vzdev
####################### [3/4]
Cleanup : httpd
####################### [4/4]
Updated: httpd.i386 0:2.0.54-10.2 vzdev.noarch 0:1.0-4.swsoft
Complete!
Updated:
httpd
i386
0:2.0.54-10.2
vzdev
noarch
0:1.0-4.swsoft
Notes:
1. Updating EZ templates is supported for running Containers only.
2. If you are going to update the cache of a commercial OS EZ template (e.g. Red Hat Enterprise Server
5 or SLES 10), you should first update software packages in the remote repository used to handle this
OS EZ template and then proceed with updating the EZ template cache. Detailed information on how to
manage repositories for commercial Linux distributions is provided in the Parallels Server Bare Metal
Templates Management Guide.
As you can see from the example above, the httpd and vzdev applications have been updated
for the redhat-el5-x86 OS EZ template. If you wish to update all EZ templates (including the
OS EZ template) inside Container 101 at once, execute this command:
# vzpkg update 101
...
Running Transaction
Updating : hwdata
Cleanup
: hwdata
###################### [1/2]
###################### [2/2]
Updated: hwdata.noarch 0:1.0-3.swsoft
Complete!
Updated:
hwdata
noarch
0:0.158.1-1
In the example above, only the hwdata package inside Container 101 was out of date and
updated to the latest version.
211
Keeping Your System Up To Date
Updating OS EZ Template Caches
With the release of new updates for the corresponding Linux distribution, the created OS EZ
template cache can become obsolete. Parallels Server Bare Metal allows you to quickly update
your OS EZ template caches using the vzpkg update cache command.
Note: If you are going to update the cache of a commercial OS EZ template (e.g. Red Hat Enterprise
Server 5 or SLES 10), you should first update software packages in the remote repository used to handle
this OS EZ template and then proceed with updating the EZ template cache. Detailed information on how
to manage repositories for commercial Linux distributions is provided in the Parallels Command Line
Reference Guide.
When executed, vzpkg update cache checks the cache directory in the template area (by
default, the template area is located in /vz/template) on the Parallels server and updates all
existing tarballs in this directory. However, you can explicitly indicate the tarball for what OS EZ
template should be updated by specifying the OS EZ template name. For example, to update the
tarball for the fedora-core-8-x86 OS EZ template, you should issue the following command:
# vzpkg
Loading
Setting
Setting
base0
base1
base2
base3
...
update cache fedora-core-8-x86
"rpm2vzrpm" plugin
up Update Process
up repositories
100% |=========================|
100% |=========================|
100% |=========================|
100% |=========================|
951
951
951
951
B
B
B
B
00:00
00:00
00:00
00:00
Upon the vzpkg update cache execution, the old tarball is renamed by receiving the -old
suffix (e.g. fedora-core-8-x86.tar.gz-old):
# ls /vz/template/cache
fedora-core-8-x86.tar.gz
fedora-core-8-x86.tar.gz-old
You can also pass the -f option to vzpkg update cache to remove an existing tar archive and
create a new one instead of it.
If the vzpkg update cache command does not find a tarball for one or several OS EZ templates
installed on the server, it creates tar archives of the corresponding OS EZ templates and puts them
to the /vz/template/cache directory.
212
CHAPTER 9
Advanced Tasks
This chapter describes those tasks that are intended for advanced system administrators who
would like to obtain deeper knowledge about Parallels Containers capabilities.
In This Chapter
Configuring Capabilities .......................................................................................... 213
Creating Customized Containers............................................................................. 218
Changing System Time From Containers ................................................................ 225
Obtaining Server ID From Inside a Container ........................................................... 226
Enabling VPN for Containers ................................................................................... 226
Managing Server Resources Parameters................................................................. 227
Setting Immutable and Append Flags for Container Files and Directories ................. 228
Customizing the /proc/meminfo Output in Containers ............................................. 229
Loading iptables Modules ....................................................................................... 231
Creating Configuration Files for New Linux Distributions .......................................... 234
Monitoring Resources............................................................................................. 235
Aligning Disks and Partitions in Virtual Machines...................................................... 236
Running Parallels Server Bare Metal 5.0 in Virtual Machines .................................... 242
Configuring Capabilities
Capabilities are sets of bits that permit of splitting the privileges typically held by the root user into a
larger set of more specific privileges. The POSIX capabilities are defined by a draft IEEE standard
(IEEE Std 1003.1e); they are not unique to Linux or Parallels Server Bare Metal. When the Linux or
Parallels Server Bare Metal documentation says “requires root privileges”, in nearly all cases it really
means “requires a specific capability”.
This section documents the tasks that can be achieved using per-Container capabilities in Parallels
Server Bare Metal and all configurable capabilities.
Advanced Tasks
Creating VZFS Symlinks Inside a Container
Normally it is impossible to create a VZFS symlink from a Container. The ability to create VZFS
symlinks presents a serious security concern explained further in this subsection. However, there
may be a situation when you need such an ability, for example, for testing created templates or
creating VZFS mounts.
A VZFS symlink is a symbolic link starting with four slashes. You can see VZFS symlinks in the
private area of any Container, as is illustrated below:
# ls -l /vz/private/101/root/bin/bash
lrwxr-xr-x
1 root root 37 Jul 9 2009 \
/vz/private/101/root/bin/bash -> \
////redhat-as4/bash-3.0-19.2/bin/bash
VZFS symlinks have no special meaning if the private area is not mounted over VZFS (to the
Container root directory). If it is, then instead of a VZFS symlink the users inside the Container will
see the file located in the template directory (in this particular case, /vz/template/redhatas4/bash-3.0-19.2/bin/bash) instead of the VZFS symlink.
If you try to create a VZFS symlink inside the Container, you will get an error:
[root@ct101 root]# ln -s ////redhat-as4/bash-3.0-19-2/bin/bash .
ln: creating symbolic link `./bash' to \
`////redhat-as4/bash-3.0-19.2/bin/bash': Invalid argument
The reason for this restriction is security considerations. If an intruder can correctly guess where the
template area (defined by the TEMPLATE variable in the global configuration file
/etc/sysconfig/vz) is located, he/she can access any file on the server provided the path to
the file is guessed correctly. However, in case it is necessary to allow the VZFS symlinks creation
inside a Container, it is possible to make use of the sys_rawio capability:
# vzctl set 101 --capability sys_rawio:on --save
Unable to set capability on running Container
Saved parameters for Container 101
After restarting the Container, you can unpack VZRPMs inside the Container or simply create VZFS
symlinks:
# ssh root@ct101
root@ct101's password:
Last login: Mon Oct 28 23:25:58 2008 from 10.100.40.18
[root@ct101 root]# rpm2cpio bash-3.0-19.2.i386.vz.rpm | cpio -id
94 blocks
[root@ct101 root]# ls -l bin/bash
-rwxr-xr-x
1 root
root
519964 Oct 29 23:35 bin/bash
[root@ct101 root]# ln -s ////redhat-as4/bash-3.0-19.2/bin/bash .
[root@ct101 root]# ls -l bash
-rwxrwxrwx
1 root
root
519964 Oct 29 23:35 bash
As you can see both VZFS symlinks look like regular files for Container users. If you need to unpack
and work on symlinks themselves, you have to create a Container that has a directory bindmounted over a regular file system such as EXT2FS, EXT3FS or ReiserFS.
214
Advanced Tasks
Remember that assigning this capability to non-trusted Containers can lead to compromising the
server. The session below shows how a malicious Container administrator can get a copy of the
server password database files:
[root@ct101 root]# ln -s ////../../etc/passwd .
[root@ct101 root]# ln -s ////../../etc/shadow .
[root@ct101 root]# ls -l
total 3
-rwxrwxrwx
1 root
root
1252 Oct 29 23:56 passwd
-rwxrwxrwx
1 root
root
823 Oct 29 23:56 shadow
While there is no easy way to substitute the password files on the server, a malicious Container
administrator could run a dictionary attack against the obtained files.
Available Capabilities for Containers
This section lists all the capabilities that can be set with the <pctl> command. The capabilities are
divided into two tables: the capabilities defined by the POSIX draft standard and Linux-specific
capabilities. For each capability, its description is given together with the default value for a
Container.
Please note that it is easy to create a non-working Container or compromise your server security by
setting capabilities incorrectly. Do not change any capability for a Container without a full
understanding of what this capability can lead to.
215
Advanced Tasks
Capabilities Defined by POSIX Draft
Name
Description
chown
If a process has this capability set on, it can change ownership on the on
files not belonging to it or belonging to another user. You have to set
this capability on to allow the Container root user to change
ownership on files and directories inside the Container.
dac_override
This capability allows to access files even if the permission is set to
disable access. Normally leave this on to let the Container root
access files even if the permission does not allow it.
dac_read_search
Overrides restrictions on reading and searching for files and
on
directories. The explanation is almost the same as above with the sole
exclusion that this capability does not override executable restrictions.
fowner
Overrides restrictions on setting the S_ISUID and S_ISGID bits on a on
file requiring that the effective user ID and effective group ID of the
process shall match the file owner ID.
fsetid
Used to decide between falling back on the old suser() or
fsuser().
on
kill
Allows sending signals to processes owned by other users.
on
setgid
Allows group ID manipulation and forged group IDs on socket
credentials passing.
on
setuid
Allows user ID manipulation and forged user IDs on socket credentials on
passing.
216
Default
on
Advanced Tasks
Linux-Specific Capabilities
Name
Description
Default
setpcap
Transfer any capability in your permitted set to any process ID;
remove any capability in your permitted set from any process ID.
off
linux_immutable
on
Allows the modification of the S_IMMUTABLE and S_APPEND file
attributes. These attributes are implemented only for the EXT2FS
and EXT3FS Linux file systems and, as such, this capability has no
effect for Containers running on top of VZFS. However, if you bind
mount a directory located on the EXT2FS or EXT3FS file system into
a Container and revoke this capability, the root user inside the
Container will not be able to delete or truncate files with these
attributes on.
net_bind_service
Allows to bind to sockets with numbers below 1024.
on
net_broadcast
Allows network broadcasting and multicast access.
on
net_admin
Allows the administration of IP firewalls and accounting.
off
net_raw
Allows to use the RAW and PACKET sockets.
on
ipc_lock
Allows to lock shared memory segments and mlock/mlockall
calls.
on
ipc_owner
Overrides IPC ownership checks.
on
sys_module
Insert and remove kernel modules. Be very careful with setting this
capability on for a Container; if a user has the permission of
inserting kernel modules, this user has essentially full control over
the server.
off
sys_rawio
Allows to create VZFS symlinks over VZFS.
off
sys_chroot
Allows to use chroot().
on
sys_ptrace
Allows to trace any process.
on
sys_pacct
Allows to configure process accounting.
on
sys_admin
In charge of many system administrator tasks such as swapping,
administering APM BIOS, and so on. Shall be set to off for
Containers.
off
sys_boot
This capability currently has no effect on the Container behaviour.
on
sys_nice
Allows to raise priority and to set priority for other processes.
on
sys_resource
Override resource limits (do not confuse with user beancounters).
on
sys_time
Allows to change the system time.
off
sys_tty_config
Allows the configuration of TTY devices.
on
mknod
Allows the privileged aspects of mknod().
on
lease
Allows to take leases of files.
on
217
Advanced Tasks
Creating Customized Containers
If you wish to run one or several customized applications inside your Containers and the number of
such Containers is relatively large, you may think of a way to automate the process of creating
Containers that already have a number of applications installed and tuned to meet your demands.
So, you do not need to manually install and customize your applications every time you create a
new Container.
Parallels Server Bare Metal allows you to create customized Containers having a certain set of
customized applications installed inside them right after their creation in one of the following ways:
•
By making a customized base OS EZ template and using it as the basis for Containers.
•
By making a non-base OS EZ template and using it as the basis for Containers.
•
By making a customized application EZ template, adding it to a new configuration sample file,
and using this sample file as the basis for Containers.
All these operations are described in the following subsections in detail.
218
Advanced Tasks
Using Customized OS EZ Templates
Let us first start with making a customized base OS EZ template which can then be used to create
Containers with a set of application already tuned to meet your demands. To make such a
template, do the following:
1
Create a metafile that will serve as the basis for your customized base OS EZ template.
Notes:
1. Detailed information on how to create metafiles is given in the Parallels Server Bare Metal
5.0Templates Management Guide.
2. While creating a metafile for a new OS EZ template, make sure that the value of either the %osname
parameter or the %version parameter in the metafile differs from the names or versions of all base OS
EZ templates installed on the server.
2
Create one or more scripts that will be executed on different stages of the OS EZ template
lifecycle and customize your applications to meet your needs. For example, you can create a
postinstall script with the name of post_install.bash and make it perform a number of
customization operations on some application included in the OS EZ template after installing
this application inside your Container.
3
Create a customized OS EZ template by running the vzmktmpl utility and passing the
corresponding options to it. So, you can use the --post-install option and specify the
path to the post_install.bash script from the example above to make an OS EZ template
that will customize your application after installing it inside your Container.
Note: The full list of options allowing you to specify what scripts are to be executed on what stage of
the EZ template lifecycle is provided in the vzmktmpl subsection of the Parallels Containers 4.6 Line
Reference Guide.
4
Install the customized OS EZ template on the server using the rpm -i command.
5
Cache the created OS EZ template by running the vzpkg create cache command.
Detailed information on how you can do it is provided in the Parallels Server Bare Metal 5.0
Templates Management Guide.
6
Create a Container based on the OS EZ template.
For example, to create a Container that will run CentOS 5 and have the customized mysql and
apache applications installed right after its creation, you can do the following:
1
Create a metafile for the Cent OS EZ template, name it, for example,
centos_5_customized.metafile, and save in the /root/centos_5 directory on the
server.
2
Make a script that will perform a number of custom operations after applying the mysql and
apache application EZ templates to the Container, and name it post_install.bash.
3
Copy the script to the /root/centos_5 directory on the server.
4
Execute the following command on the server to create the CentOS 5 OS EZ template:
219
Advanced Tasks
# vzmktmpl /root/centos_5/centos_5_customized.metafile \
--post-install /root/centos5/post_install.bash
This command will create an OS EZ template for CentOS and put it to the /root directory (for
example, /root/centos_customized-5-x86-ez-4.7.0-1.noarch.rpm).
5
Install the resulting OS EZ template on the server:
# rpm -i /root/centos_customized-5-x86-ez-4.7.0-1.noarch.rpm
6
Cache the installed OS EZ template:
# vzpkg create cache centos_customized-5-x86
...
Complete!
Packing cache file centos_customized-5-x86.tar.gz ...
Cache file centos_customized-5-x86.tar.gz [14M] created.
7
Create Container 101 on the basis of the new OS EZ template:
# pctl create 101 --ostemplate centos_customized-5-x86
-–config basic
Creating Container private area (centos_customized-5-x86)
Container is mounted
Postcreate action done
Container is unmounted
Container private area was created
Delete port redirection
Adding port redirection to Container(1): 4643 8443
So you have just created Container 101 having the customized mysql and apache applications
installed inside it.
220
Advanced Tasks
Using EZ OS Template Sets
Another way of creating customized Containers is to make a non-base OS EZ template (also
known as an OS EZ template set) differing from the corresponding base OS EZ template in the
number of packages included in this template. For example, if you wish a Container to run CentOS
5 and to function as a Linux-based server only, you can create the centos-5-x86-server OS
EZ template set and include only those packages in it that are needed for performing main server
tasks. So, you can specify packages to be used for setting up file and print sharing and exclude all
the packages for graphical interfaces (GNOME and KDE).
To create a non-base OS EZ template, do the following:
1
Create a metafile that will serve as the basis for your non-base OS EZ template. Any metafile
for this kind of EZ template must contain the following information:
• %osname: the name of the Linux distribution for which you are creating the OS EZ template
set. This name must correspond to that specified in the base OS EZ template. For example,
if you are creating an OS template set of the base OS EZ template for CentOS 5, set the
value of this parameter to centos.
• %osver: the version of the Linux distribution specified as the value of the %osname
parameter. This name must correspond to that specified in the base OS EZ template. For
example, if you are creating an OS template set of the base OS EZ template for CentOS 5,
set the value of this parameter to 5.
• %osarch: the system architecture where the EZ template is to be run. This name must
correspond to that specified in the base OS EZ template. For example, if you are creating an
OS template set of the base OS EZ template for CentOS 5, set the value of this parameter
to x86.
• %setname: the name to be assigned to your non-base OS EZ template. You can specify
any name you like for your OS template set:
a
This name will be added to the name of the base OS EZ template after the indication of the
architecture where the OS EZ template is to be run. For example, if you are creating an OS
template set of the base OS EZ template for CentOS 5 that is supposed to run on x86
platforms, the name of your non-base OS EZ template should look like the following—
centos-5-x86-Template_Name-ez-1.0-1.noarch.rpm—where Template_Name
is the name you specify as the value of the %setname parameter.
b
This name will also be assigned to the directory which will store the meta data of your nonbase OS EZ template after the template installation on the server. For example, it will have
the name of /vz/template/centos/5/x86/config/os/my_non_base_template
if you set the value of this parameter to my_non_base_template, create a non-base OS
EZ template for CentOS 5, and installed it on the server.
221
Advanced Tasks
• %packages: a list of RPM packages to be included in the non-base OS EZ template. This
parameter allows you to specify what applications will be present inside your Containers
based on this OS EZ template set right after their installation. The names of the packages
listed as the value of this parameter must correspond to the names of real RPM packages
(without indicating the package version, release, architecture, and the .rpm extension) that
are stored in the repository used for managing your EZ templates.
Note: You can also specify a number of additional parameters in your metafile. For example, you may
wish to add one or several extra packages to your OS EZ template set which are not available in the
repository used to handle the packages for the corresponding base OS EZ template. For this purpose,
you will have to specify the %mirrorlist parameter providing information on the repository where
these extra packages are kept. Detailed information on all parameters you can set in metafiles is given in
the Parallels Containers 4.6 Reference Guide.
2
You can also (though you do not have to) create a number of scripts that will be executed on
different stages of the non-base OS EZ template lifecycle and customize your applications to
meet your demands. The path to these scripts should then be specified after the corresponding
options while creating your OS template set. For example, you can create a preinstall script with
the name of pre_install.bash and make it perform a number of customization operations
on some application included in the non-base OS EZ template before installing this application
in your Container.
Note: If there are no scripts for a non-base OS EZ template, the scripts available for the corresponding
base OS EZ template will be executed.
3
Create the non-base OS EZ template by running the vzmktmpl utility and passing the
corresponding options to it, if needed. So, if you created one or several scripts in the previous
step, you can use special options and specify the path to these scripts during the command
execution. For example, you can use the --pre-install option and specify the path to the
pre_install.bash script to make an OS EZ template that will customize your application
before installing it inside your Container.
Note: The full list of options allowing you to specify what scripts are to be executed on what stage of
the EZ template lifecycle is provided in the vzmktmpl subsection of the Parallels Containers 4.6
Reference Guide.
4
Install the non-base OS EZ template on the server using the rpm -i command.
5
Cache the created OS EZ template by running the vzpkg create cache command.
Detailed information on how you can do it is provided in the Parallels Server Bare Metal 5.0
Templates Management Guide.
6
Create a Container based on the OS EZ template.
222
Advanced Tasks
Using Customized Application Template
If the number of customized applications inside your Containers is relatively small, you can also use
the following way of creating customized Containers:
1
Create a metafile that will serve as the basis for your customized application EZ template.
Note: Detailed information on how to create metafiles is given in the Creating Metafiles for EZ
Templates section of the Parallels Server Bare Metal 5.0 Templates Management Guide.
2
Create one or more scripts that will be executed on different stages of the application EZ
template lifecycle and customize your applications to meet your demands. For example, you
can create a postinstall script with the name of post_install.bash and make it perform a
number of customization operations on your application after installing this application in your
Container.
3
Create a customized application EZ template by running the vzmktmpl utility and passing the
corresponding options to it. So, you can use the --post-install option and specify the
path to the post_install.bash script from the example above to customize your
application in accordance with your needs after installing it in your Container.
Note: The full list of options allowing you to specify what scripts are to be executed on what stage of
the EZ template lifecycle is provided in the vzmktmpl section of the Parallels Containers 4.6 Reference
Guide.
4
Install the customized EZ template on the server using the rpm -i command.
5
Create a new Container configuration sample file and include the customized EZ template in
this file. Detailed information on Container configuration sample files is provided in the
Managing Container Resources Configuration section (p. 149).
6
Create a customized Container on the basis of the configuration sample.
The following example demonstrates how to create Container 101 that will run CentOS 5 and have
the customized mysql application installed right after its creation:
1
Create a metafile for the mysql application, name it mysql.metafile, and save in the
/usr/mysql directory on the server.
2
Make a script that will perform a number of custom operations after applying the mysql EZ
template to the Container, and name it post_install.bash.
3
Copy the script to the /usr/mysql directory on the server.
4
Execute the following command on the server to create the mysql EZ template:
# vzmktmpl /usr/mysql/mysql.metafile \
--post-install /usr/mysql/post_install.bash
This command will create an EZ template for the mysql application and put it to the /root
directory (e.g., /root/mysql-centos-5-x86-ez-4.0.0-17.swsoft.noarch.rpm).
5
Install the mysql EZ template on the server. Using the example above, you can install the
template as follows:
223
Advanced Tasks
# rpm -ihv /root/mysql-centos-5-x86-ez-4.0.0-17.swsoft.noarch.rpm
6
Create a new Container configuration sample file and add the mysql EZ template to a list of
templates that will be installed in Containers created on the basis of this configuration sample
file.
7
Create Container 101 by using the pctl create command and the mysql sample file:
# pctl create 101 --ostemplate centos-5-x86 -–config mysql
Creating Container private area (centos-5-x86)
Container is mounted
Postcreate action done
Container is unmounted
Container private area was created
Delete port redirection
Adding port redirection to Container(1): 4643 8443
So, you have just created Container 101 that already has the customized mysql application
installed.
224
Advanced Tasks
Changing System Time From Containers
Normally, it is impossible to change the system time from a Container. Otherwise, different
Containers could interfere with each other and could even break applications depending on the
system time accuracy.
Normally only the server system administrator can change the system time. However, if you want to
synchronize the time via Network Time Protocol (NTP), you have to run NTP software, which will
connect to external NTP servers and update the system time. It is not advisable to run application
software on the server itself, since flaws in the software can lead to compromising all Containers on
this server. Thus, if you plan to use NTP, you should create a special Container for it and configure
it to have the sys_time capability. The example below illustrates configuring such a Container:
# pctl set 101 --capability sys_time:on --save
Unable to set capability on running Container
Saved parameters for Container 101
The output of the above command warns you that pctl cannot apply changes in the capabilities
to a running Container. The Container has to be restarted before changes take effect:
# pctl stop 101; pctl start 101
Stopping Container ...
Container was stopped
Container is unmounted
Starting Container ...
Container is mounted
Adding IP address(es): 192.168.1.101
Hostname for Container set: Container101
Container start in progress...
# ssh root@ct101
root@ct101's password:
Last login: Mon Feb 28 23:25:58 2007 from 10.100.40.18
[root@ct101 root]# date
Mon Feb 28 23:31:57 EST 2007
[root@ct101 root]# date 10291300
Tue Feb 29 13:00:00 EST 2007
[root@ct101 root]# date
Tue Feb 29 13:00:02 EST 2007
[root@ct101 root]# logout
Connection to Container101 closed.
# date
Tue Feb 29 13:01:31 EST 2010
The command session above shows the way to change the system time from Container 101. The
changes will affect all the Containers and the server itself. It is not advisable to have more than one
Container with the sys_time capability set on.
NTP is described in Internet Standard RFC 1305; more information including client software can be
obtained from the NTP web server (http://www.ntp.org).
225
Advanced Tasks
Obtaining Server ID From Inside a Container
The default Parallels Server Bare Metal installation does not allow users inside a Container to obtain
any information specific to the Parallels server the Container is running on. The reason is that no
Container shall have knowledge about the corresponding server. A Container can be transparently
migrated to another server, and if this Container runs any applications depending on the particular
server, these applications might fail after the migration.
In some situations, however, you need to provide a unique server ID to some applications. For
example, you might want to license your application per server. In this case, after the migration your
customer will need to re-apply the license for your application.
Parallels Server Bare Metal provides access to the unique server ID via the /proc/vz/hwid file.
The default Parallels Server Bare Metal installation makes this file accessible to Containers from 1 to
100 (i.e. Containers with reserved IDs). It is possible to change this range in the global configuration
file (vz.conf). For example, this is the way to make the file visible in Containers from 1 to 1000:
# vi /etc/vz/vz.conf
VZPRIVRANGE=”1 1000”
# pctl exec 101 cat /proc/vz/hwid
0C3A.14CB.391B.6B69.02C9.4022.3E2F.CAF6
The above example illustrates accessing the server ID from Container 101.
Enabling VPN for Containers
Virtual Private Network (VPN) is a technology which allows you to establish a secure network
connection even over an insecure public network. Setting up a VPN for a separate Container is
possible via the TUN/TAP device. To allow a particular Container to use this device, the following
steps are required:
•
Make sure the tun.o module is already loaded before Parallels Server Bare Metal is started:
# lsmod
•
Allow the Container to use the TUN/TAP device:
# pctl set 101 --devices c:10:200:rw --save
•
Create the corresponding device inside the Container and set the proper permissions:
# pctl exec 101 mkdir -p /dev/net
# pctl exec 101 mknod /dev/net/tun c 10 200
# pctl exec 101 chmod 600 /dev/net/tun
Configuring the VPN proper is carried out as a common Linux administration task, which is out of
the scope of this guide. Some popular Linux software for setting up a VPN over the TUN/TAP driver
includes Virtual TUNnel <http://vtun.sourceforge.net/> and OpenVPN
<http://openvpn.sourceforge.net/>.
226
Advanced Tasks
Managing Server Resources Parameters
Parallels Server Bare Metal allows you to configure a number of resource management parameters
defining the amount of resources to be allocated to the Parallels server. These parameters include
all standard UBC parameters (VMGUARPAGES, KMEMSIZE, OOMGUARPAGES, etc.) as well as the
ONBOOT parameter.
You can edit any of these parameters in the /etc/vz/conf/0.conf file on the server using your
favorite text editor (for example, vi or emacs) or using the pctl set command and specifying 0
after this command. For example:
# pctl set 0 --kmemsize 12211840:14359296 --save
Saved parameters for Container 0
This command sets both the barrier and limit values of unswappable kernel memory (in bytes)
which can be allocated to internal kernel structures of the processes on the server. The specified
parameter values will be in force until the server restart. If you wish these values to be applied to the
server on its next booting, you should additionally set the ONBOOT parameter in the
/etc/vz/conf/0.conf file to yes. This can be done in one of the following ways:
•
Passing the --onboot option to the pctl set command:
# pctl set 0 --onboot yes
Saved parameters for Container 0
•
Editing the /etc/vz/conf/0.conf file with your favorite text editor (e.g. vi) and setting the
value of the ONBOOT parameter in this file to yes.
Note: Detailed information on all resource parameters that can be changed in respect of your Parallels
server is provided in the Parallels Command Line Reference Guide.
If you have made a number of changes to server resource management parameters and wish to
reset them to the values specified in the /etc/vz/conf/0.conf file, you can run this command:
# pctl set 0 --reset_ub
UBC limits were set successfully
227
Advanced Tasks
Setting Immutable and Append Flags for
Container Files and Directories
You can use standard Linux utilities—chattr and lsattr—to set extra flags for files and
directories inside your Containers and to query their status, respectively. Currently, two of these
extra flags—'append' and 'immutable'—are supported. For example, you can execute the following
command to set the 'immutable' flag for the /root/MyFile file inside Container 101:
[root@ct101 root] chattr +i /root/MyFile
To check that the 'immutable' flag has been successfully set, use the following command:
[root@ct101 root] lsattr /root/MyFile
----i-------- /root/MyFile
Note: For detailed information on the chattr and lsattr utilities, see their manual pages.
228
Advanced Tasks
Customizing the /proc/meminfo Output in
Containers
The /proc/meminfo virtual file allows you to view the information about memory usage (both
physical and swap) on the system.You can customize the output of this file inside a particular
Container and set it to one of the following modes:
•
Non-virtualized. In this case running the cat /proc/meminfo command inside a Container
will display the information about the physical memory on the server (total, used, free, shared,
etc.), in kilobytes.
•
Virtualized in pages. Setting the /proc/meminfo output to this mode allows you to specify
what amount of total memory (in kilobytes) will be displayed while running the cat
/proc/meminfo command inside this or that Container.
•
Virtualized in privvmpages. Setting the /proc/meminfo output to this mode also allows you to
arbitrarily specify the amount of total memory (in kilobytes) to be displayed while running the
cat /proc/meminfo command inside this or that Container. As distinct from the previous
mode, the amount of memory to be shown in this mode is calculated on the basis of the value
of the PRIVVMPAGES parameter set in the Container configuration file.
Notes:
1. In Parallels Containers 4.6, the functionality of customizing the /proc/meminfo output is obsolete.
2. Enabling a specific mode for a Container does not influence the real resources allocation to the
Container. It is only used to modify the way the /proc/meminfo output is displayed in the Container.
During the Parallels Server Bare Metal installation, the output of the /proc/meminfo virtual file is
set to the 'non-virtualized' mode, i.e. running the cat /proc/meminfo command inside any
Container will show the information about the memory usage on the Parallels server. You can use
the --meminfo option with the pctl set command to switch between different modes:
•
To set the output of /proc/meminfo inside Container 101 to the 'virtualized in pages' mode,
issue the following command:
# pctl set 101 --meminfo pages:2000 --save
The amount of memory that will be displayed by running the cat /proc/meminfo command
inside Container 101 is defined by the data specified after the --meminfo option:
• pages tells the pctl set command to enable the 'virtualized in pages' mode for the
/proc/meminfo output and simultaneously denotes the units of measurement to be used
for setting the amount of memory (e.g., 4-KB pages for Containers running 32-bit operating
systems).
• 200 denotes the number of pages to be shown in the /proc/meminfo output.
In our case the /proc/meminfo output inside Container 101 may look like the following:
229
Advanced Tasks
# pctl exec 101 cat /proc/meminfo
MemTotal:
8000 kB
MemFree:
5140 kB
LowTotal:
8000 kB
LowFree:
5140 kB
Buffers:
0 kB
Cached:
0 kB
SwapCached:
0 kB
HighTotal:
0 kB
HighFree:
0 kB
...
When working in this mode, keep in mind the following:
• The specified amount of memory (in our case it is 8000 KB) is always shown in the
MemTotal and LowTotal fields of the cat /proc/meminfo output.
• The values in the MemFree and LowFree fields are calculated automatically by the system.
• All the other fields in the command output have the values set to 0.
•
To set the output of /proc/meminfo inside Container 101 to the 'virtualized in privvmpages'
mode, execute the following command:
# pctl set 101 --meminfo privvmpages:3 --save
The amount of memory that will be displayed by running the cat /proc/meminfo command
inside Container 101 is calculated using the following formulas:
• Privvmpages_Value * 3 * 4KB if Container 101 is running a 32-bit operating system
(OS) or an OS for x86-64 processors and
• Privvmpages_Value * 3 * 16KB if Container 101 is running an OS for IA-64
processors
where Privvmpages_Value denotes the value of the PRIVVMPAGES parameter set in the
Container configuration file and 3 is an arbitrary integer coefficient which you can modify to
increase/decrease the amount of memory in the /proc/meminfo output. Assuming that the
privvmpages parameter for Container 101 is set to 10000, your output may look as follows:
# pctl exec 101 cat /proc/meminfo
MemTotal:
120000 kB
MemFree:
78248 kB
LowTotal:
120000 kB
LowFree:
78248 kB
Buffers:
0 kB
Cached:
0 kB
SwapCached:
0 kB
HighTotal:
0 kB
HighFree:
0 kB
...
As can be seen from the example above, the displayed records comply with the same rules as
the records in the 'virtualized in pages' mode.
•
To revert the output of /proc/meminfo to the default mode, execute the following command:
# pctl set 101 --meminfo none --save
Note: If the value specified after the --meminfo option exceeds the total amount of memory available
on the Parallels server, the cat /proc/meminfo command executed inside a Container will display the
information about the total physical memory on this server.
230
Advanced Tasks
The --save flag in the commands above saves all the parameters to the Container configuration
file. If you do not want the applied changes to persist, you can omit the --save option and the
applied changes will be valid only till the Container shutdown.
Loading iptables Modules
The given section provides information on how you can manage iptables modules on the server
and in Containers.
Loading iptables Modules to Parallels Server
You can configure a list of iptables modules that will be loaded on the Parallels server after its
startup by editing the /etc/vz/vz.conf file. The IPTABLES parameter in this file determines the
iptables modules that will additionally be loaded to the server during the Parallels Server Bare
Metal startup. For example, you can indicate the following iptables modules as the value of this
parameter to have them automatically loaded to your Parallels server after the Parallels Server Bare
Metal startup:
IPTABLES="ipt_REJECT ipt_tos ipt_limit ipt_multiport iptable_filter
iptable_mangle ipt_TCPMSS
ipt_tcpmss ipt_ttl ipt_length"
All the specified modules will be loaded on the server startup once you restart it.
231
Advanced Tasks
Loading iptables Modules to Containers
The list of iptables modules that are loaded to a Container by default is determined by the
iptables modules loaded on the server at the moment of the Container startup. For example, if
your server has the ipt_REJECT, ipt_tos, ipt_limit, ipt_multiport, and
iptable_filter modules loaded, any Containers on this server will also have these iptables
modules loaded after their startup.
However, Parallels Server Bare Metal allows you to prevent certain modules from being loaded
inside a Container on its startup, even if they are loaded on the server itself. The full list of such
iptables modules is listed below:
•
ip_table
•
ip6_table
•
iptable_filter
•
ip6table_filter
•
iptable_mangle
•
ip6table_mangle
•
ip_conntrack
•
ip_conntrack_ftp
•
ip_conntrack_irc
•
iptable_nat
•
ip_nat_ftp
•
ip_nat_irc
To forbid the usage of any of the aforementioned iptables modules inside a Container, you
should explicitly indicate the names of the modules you wish to be loaded to the Container as the
value of the IPTABLES parameter in the Container configuration file
(/etc/vz/conf/<CT_ID>.conf) or by using the pctl command. For example:
# pctl set 101 --iptables ip_table --iptables iptable_filter --iptables ip_conntrack -iptables iptable_nat --iptables iptable_mangle --save
This command will tell Parallels Server Bare Metal to:
•
load the ip_table, iptable_filter, ip_conntrack, iptable_nat, and
iptable_mangle modules to Container 101 if they are loaded on the server during the
Container startup
•
forbid the usage of all the other iptables modules listed above (i.e. ip6_table,
ip6table_filter, ip6table_mangle, ip_conntrack_ftp, ip_conntrack_irc,
ip_nat_ftp, ip_nat_irc) inside Container 101 even if they are loaded on the server during
the Container startup
232
Advanced Tasks
This information will also be saved in the Container configuration file thanks to the --save option.
Loading a new set of iptables modules does not happen on the fly. You must restart the
Container for the changes to take effect.
233
Advanced Tasks
Creating Configuration Files for New Linux
Distributions
Distribution configuration files are used to distinguish among Containers running different Linux
versions and to determine what scripts should be executed when performing the relevant
Container-related operations (e.g. assigning a new IP address to the Container). Detailed
information on distributions configurations files is provided in the Linux Distribution Configuration
Files subsection of the Parallels Command Line Reference Guide.
All Linux distributions shipped with Parallels Server Bare Metal have their own configuration files
located in the /etc/vz/conf/dists directory on the Parallels server. However, you may wish to
create your own distribution configuration files to support new Linux versions released. Let us
assume that you wish your Containers to run the CentOS 5 Linux distribution and, therefore, have
to make the centos-5.conf distribution configuration file to define what scripts are to be
executed while performing major tasks with Containers running this Linux version. To do this:
1
In the Container configuration file (with the name of /etc/vz/conf/CT_ID.conf), specify
centos-5 as the value of the DISTRIBUTION variable (for example,
DISTRIBUTION="centos-5").
2
Create the centos-5.conf configuration file in the /etc/vz/conf/dists directory. The
easiest way to do it is copy one of the existing configuration files by executing the following
command in the /etc/vz/conf/dists directory:
# cp fedora.conf centos-5.config
In the example above, we assume that the fedora.conf file is present in the
/etc/vz/conf/dists directory on the Parallels server. In case it is not, you may use any
other distribution configuration file available on your server.
3
Open the centos.conf file for editing with the help of any text editor:
# vi centos-5.conf
4
In the centos-5.conf file, go to the first entry and, in the right part of the entry, specify the
name of the script you wish to be run on issuing the pctl command with the parameter
specified in the left part of the entry. For example, if you wish the script to be executed while
assigning a new IP address to your Container and the script has the my_centos_script
name, your entry should look as follows:
ADD_IP=my_centos_script-add_ip.sh
Note: The information on all acceptable parameters and their description are provided in the Parallels
Command Line Reference Guide.
5
Repeat Step 4 for all entries in the file.
6
Place the scripts for the new Linux distribution to the /etc/vz/conf/dists/scripts
directory on the Node. Make sure the names of these scripts coincide with those specified in
the centos-5.conf file.
234
Advanced Tasks
Monitoring Resources
In Parallels Server Bare Metal, you can use the pstat utility to monitor system resources in real
time. When executed, the utility displays the status and load of the system: its disk, network, CPU,
memory, and other parameters. It also provides the list of running virtual machines and Containers
together with their resources consumption statistics. For example, you can run the following
command on the server to view your current system resources:
# pstat -d 5
5:39pm, up 4 days, 5:33, 2 users, load average: 1.08, 1.11, 1.05
CTNum 2, procs 268: R
1, S 265, D
1, Z
0, T
1, X
0
CPU [ OK ]: CTs
0%, CT0
0%, user
0%, sys
1%, idle 99%, lat(ms)
1/0
Mem [ OK ]: total 7831MB, free 4147MB/0MB (low/high), lat(ms) 1/0
ZONE0 (DMA): size 9MB, act 0MB, inact 0MB, free 10MB (0/0/0)
ZONE1 (DMA32): size 3238MB, act 42MB, inact 39MB, free 3118MB (4/5/6)
ZONE2 (Normal): size 4661MB, act 2730MB, inact 606MB, free 1019MB (6/8/9)
Mem lat (ms): A0 0, K0 1, U0 0, K1 1, U1 0
Slab pages: 181MB/181MB (ino 39MB, de 13MB, bh 21MB, pb 40MB)
Swap [ OK ]: tot 2000MB, free 2000MB, in 0.000MB/s, out 0.000MB/s
Net [ OK ]: tot: in 0.027MB/s 233pkt/s, out 0.040MB/s
37pkt/s
lo: in 0.000MB/s
0pkt/s, out 0.000MB/s
0pkt/s
eth0: in 0.014MB/s 116pkt/s, out 0.020MB/s
19pkt/s
sit0: in 0.000MB/s
0pkt/s, out 0.000MB/s
0pkt/s
br0: in 0.000MB/s
0pkt/s, out 0.000MB/s
0pkt/s
br1: in 0.013MB/s 116pkt/s, out 0.020MB/s
19pkt/s
Disks [ OK ]: in 0.000MB/s, out 0.000MB/s
ST
%VM
OK 0.0/27
OK 0.2/685
OK 0.4/685
OK 27/6.7
%KM
0.0/0.0/0.0/0.1/-
CPU
0.00/33
0.00/33
0.00/33
0.00/33
FCNT MLAT NAME
0
0 1
0
0 101
0
0 102
0
0 Windows7
The command output is updated with the time interval equal to the value specified after the –d
(delay) option measured in seconds. In the session above, the statistics displayed is renewed every
five seconds. If the –d option is not specified, the default interval equals 1 second.
As you can see, the utility provides real-time information on all main resources subsystems
pertaining both to the physical server and to its virtual machines and Containers: the disk, network,
CPU, and memory subsystems. You may want to shrink the output of the utility by specifying the –
b (brief) option instead of the –v (verbose) one, or to do without any options to use the “normal”
mode of displaying.
The following information is displayed by default per each virtual machine or Container:
Column Name
Description
ST
virtual machine or Container status. If there are no failed counters and the latency
values are normal, the status is “OK”. Otherwise, it is displayed in red as “!!”. You can
sort virtual machines and Containers by their status to see the problem virtual
machines and Containers first.
%VM
Virtual memory usage, in per cent to the total memory. The first number is how much
virtual memory is being used, and the second one is the virtual memory barrier.
235
Advanced Tasks
%KM
Kernel memory usage, in per cent to the normal zone size. The first number is how
much kernel memory is being used, and the second one is the kernel memory barrier.
CPU
CPU usage in per cent to all available CPUs. The first number is how much of the CPU
power is being used by the virtual machine or Container, and the second one is its
guaranteed share judging by the cpuunits parameter. Note that the actual CPU
usage may be higher than the guaranteed one.
FCNT
The number of failed counters for all the resource parameters. In the standard mode
of displaying, this number represents the increase of failed counters since the previous
screen update, whereas in the average mode of displaying, it represents an absolute
failed counters sum for the given virtual machine or Container.
MLAT
Maximal scheduling latency for the virtual machine or Container, in ms. This parameter
shows the maximal scheduling latency inside the given virtual machine or Container,
i.e. for how long (at the utmost) a process inside the virtual machine or Container
awaits for the CPU.
NAME
virtual machine or Container name.
The %VM, %KM, and CPU columns provide two values per column separated by a slash for each
virtual machine and Container. The first value indicates the real usage of the corresponding
parameter by the virtual machine and Container, and the second one – the maximal value allowed
for the virtual machine and Container.
For detailed information on options that you can use with the pstat utility, refer to the Parallels
Server Bare Metal Command Linux Guide.
Aligning Disks and Partitions in Virtual Machines
Most of the modern operating systems (Windows Server 2008, Red Hat Enterprise Linux 6, or
Fedora 14) automatically align the partitions when you install them in virtual machines. For example,
Windows Server 2008 creates a default partition offset of 1024 KB to satisfy the partition alignment
requirements. The following figure shows an example of the correct partition alignment:
236
Advanced Tasks
In this example, any cluster (the smallest unit of data) in the guest OS file system is aligned with the
boundaries of an NFS block, and reading from or writing to a cluster requires only access to one
NFS block. For example, reading from Cluster 1 causes only a read from Block 1.
At the same time, virtual machines running non-modern systems (for example, Windows Server
2003 or Red Hat Enterprise Linux 5) do usually have misaligned partitions, which is shown in the
figure below:
In this example, clusters of the guest OS file system do not match the boundaries of NFS blocks,
and reading from or writing to a cluster requires access to several NFS blocks. For example,
reading from Cluster 1 causes two reads: from Block 1 and from Block 2. This results in a slower
read time as compared to properly aligned partitions and leads to performance degradation.
Aligning partitions
Basically, to align disks and partitions in virtual machines, you need to set an offset so that clusters
in the guest OS file system match the volume block size on your NFS storage. Usually, the block
size of most network storages is 512 bytes or a multiple of 512 bytes. As an example, the following
sections describe the procedure of aligning disks and partitions for Linux and Windows virtual
machines assuming that the size of your NFS blocks is 512 bytes.
When deciding on aligning disks and partitions, take into account that this process destroys all data
on these disks and partitions. So if you want to have a correctly aligned system partition, you need
to align your disks and partitions before creating a virtual machine and installing a guest operating
system in it. If you do not want an aligned system partition, you can first create a virtual machine
and install a guest OS in it, and then align your data disks from inside the virtual machine.
The sections below demonstrate how to align disks and partitions before you start installing a guest
OS. You can, however, use a similar procedure to align data disks and partitions from inside your
virtual machines.
237
Advanced Tasks
Checking partition alignment in existing virtual machines
First of all, you may wish to know how you can check that the partitions of a virtual machine are not
aligned. Depending on the operating system installed in the virtual machine, you can do the
following.
Linux virtual machines
To check the partition alignment in a Linux virtual machine, log in to this virtual machine and run the
following command:
# fdisk -l -u /dev/device_name
For example, to check the partition alignment on the sdc device, you can run this command:
# fdisk -l -u /dev/sdc
Disk /dev/sdc: 73.0 GB, 73014444032 bytes
255 heads, 63 sectors/track, 8876 cylinders, total 142606336 sectors
Units = sectors of 1 * 512 = 512 bytes
Device Boot
/dev/sdc1
*
/dev/sdc2
Start
63
208845
End
208844
142592939
Blocks
104391
71192047+
Id
83
8e
System
Linux
Linux LVM
Pay attention to the number of sectors in the Start column. Usually, a sector contains 512 bytes,
which makes up 32256 bytes for 63 sectors for the /dev/sdc1 partition and 26105625 bytes for
208845 for the /dev/sdc2 partition. For a partition to be properly aligned, it must align with 4096
byte boundaries (assuming that the block size of your storage is 4 KB). As 32256 and 106928640
is not a multiple of 4096, the partitions /dev/sdc1 and /dev/sdc2 are not aligned properly. To
align them, you should offset
•
the /dev/sdc1 partition by 1 sector so that it starts at 64. In this case, 64 sectors each
containing 512 bytes make up 32768 that is a multiple of 4096.
•
the /dev/sdc2 partition by 3 sectors so that it starts at 208848. In this case, 208848 sectors
each containing 512 bytes make up 106930176 that is a multiple of 4096.
Windows virtual machines
To check the partition alignment in a Windows virtual machine, do the following:
1
Click Start > Run, type msinfo32.exe, and press Enter to open System Information.
2
Navigate to Components > Storage > Disks, and look for the Partition Starting Offset field in
the right part of the window.
238
Advanced Tasks
To find out if the partition is aligned properly, use the method described above for Linux virtual
machines.
Aligning disks for Linux virtual machines
To align partitions for use in a Linux virtual machine, you need a working Linux virtual machine.
Once you have it at hand, follow the steps below:
1
Create a new disk for the virtual machine.
On this disk, you will create aligned partitions. Then you will connect the disk to a new virtual
machine and install your Linux guest OS on this disk.
2
Start the virtual machine and log in to it using SSH.
3
Run the fdisk utility for the disk you want to align.
4
Create a primary partition, and set the starting block number for the created partition.
5
Repeat steps 4-6 to create and align all partitions you plan to have in your new virtual machine.
The following example creates partition #1 with the size of 1 GB on the /dev/sda device and uses
the offset of 64 KB.
# fdisk /dev/sda
Device contains neither a valid DOS partition table, nor Sun, SGI or OSF disklabel
Building a new DOS disklabel. Changes will remain in memory only,
until you decide to write them. After that, of course, the previous
content won't be recoverable.
239
Advanced Tasks
The number of cylinders for this disk is set to 1044.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)
Warning: invalid flag 0x0000 of partition table 4 will be corrected by w(rite)
Command (m for help): n
Command action
e
extended
p
primary partition (1-4)
p
Partition number (1-4): 1
First sector (63-16777215, default 63): 64
Last sector or +size or +sizeM or +sizeK (64-16777215, default 16777215): 208848
Command (m for help): w
The partition table has been altered!
Calling ioctl() to re-read partition table.
Syncing disks.
Once you align all the necessary partitions, disconnect the disk from the virtual machine. When
creating a new virtual machine, choose this disk for use with this virtual machine.
Aligning partitions for Windows virtual machines
To align a disk for a Windows virtual machine, you need a working Windows virtual machine. Once
you have it at hand, you can use the diskpart or diskpar utility (depending on your operating
system) to align the disk:
1
Create a new disk for the virtual machine.
On this disk, you will create aligned partitions. Then you will connect the disk to a new virtual
machine and install your Windows guest OS on this disk.
2
Open the command-line prompt, and run the diskpart or diskpar utility.
3
Select the disk you want to align.
4
Create the primary partition on the disk, and align it.
5
Exit the diskpart utility, and close the command-line prompt.
The following example demonstrates how to use the diskpart utility to align disk 1 by setting the
offset of 64 for it:
240
Advanced Tasks
Once you align the virtual disk, disconnect it from the virtual machine. When creating a new virtual
machine, choose this disk for use with this virtual machine.
Creating a template of a virtual machine with aligned partitions
To facilitate the procedure of creating virtual machines that have aligned system partitions, you can
create a template of the aligned virtual machine and deploy new virtual machines from this
template.
For example, if you align a disk by following the steps in Aligning partitions for Windows virtual
machines, then create a new virtual machine that uses this disk, and then install Windows Server
2003 operating system in the virtual machine, you will have a clean Windows Server 2003
installation on the correctly aligned disk. Now you can create a template of this virtual machine and
use this template each time you need to deploy a new virtual machine with Windows Server 2003.
241
Advanced Tasks
Running Parallels Server Bare Metal 5.0 in Virtual
Machines
Starting with version 5, you can install Parallels Server Bare Metal in virtual machines. Running
Parallels Server Bare Metal 5.0 in a virtual machine may prove useful if you want to evaluate the
product but do not have a spare physical server.
To run virtual machines with Parallels Server Bare Metal 5.0, a physical server must meet the
following requirements:
Have the following architecture:
•
Intel with VT-x and EPT (Nehalem, Westmere, SandyBridge)
•
AMD SVM and RVI (Barcelona, MagnyCore, Bulldozer)
Run one of the following products:
•
Parallels Desktop for Mac 7
•
Parallels Workstation for Windows 6
•
Parallels Workstation for Linux 6
•
VMware Fusion 3
Recommended Virtual Machine Configuration
A virtual machine is best optimized for use with Parallels Server Bare Metal 5.0 if it has the following
configuration:
•
CPU:
2 or more virtual CPUs
•
Memory:
2 GB or more RAM
•
Hard disk:
40 GB or more disk space; the disk must be of the plain type.
The process of installing Parallels Server Bare Metal in a virtual machine does not differ from that
you use to install the product on a standalone server. For detailed installation instructions, refer to
the Parallels Server Bare Metal 5.0 Installation Guide.
Restrictions and Peculiarities
When using Parallels Server Bare Metal in a virtualized environment, keep in mind the following
restrictions and specifics:
242
Advanced Tasks
•
Running Parallels Server Bare Metal in a virtual machine is intended for evaluation purposes
only. You are not recommended to use such installations in production.
•
If you change the configuration of a virtual machine where Parallels Server Bare Metal is
installed, you may need to reactivate the product.
•
VMware Fusion may show a warning when you start a virtual machine with Parallels Server Bare
Metal that it requires full access to the network traffic. Ignore this message, and proceed with
booting the virtual machine.
•
To run in a virtualized Parallels Server Bare Metal environment, a virtual machine must have
Parallels Tools installed. So before starting a virtual machine for the first time, make sure that
you have installed Parallels Tools in it.
•
Virtual machines may run only 32-bit operating systems and have only one CPU.
243
CHAPTER 10
Troubleshooting
This chapter provides the information about those problems that may occur during your work with
Parallels Server Bare Metal and suggests the ways to solve them, including getting technical
support from Parallels.
In This Chapter
General Considerations .......................................................................................... 245
Kernel Troubleshooting ........................................................................................... 247
Problems With Container Management ................................................................... 249
Getting Technical Support ...................................................................................... 252
Troubleshooting
General Considerations
The general issues to take into consideration when troubleshooting your system are listed below.
You should read them carefully before trying to solve more specific problems.
•
Make sure a valid license is always loaded on the server. If your license has expired and the
grace period is over, all the virtual machines and Containers on your server will be stopped.
•
You should always remember where you are currently located in your terminal. Check it
periodically using the pwd, hostname, ifconfig, cat /proc/vz/veinfo commands.
One and the same command executed inside a virtual machine and Container and on the
server can lead to very different results. You can also set up the PS1 environment variable to
show the full path in the bash prompt. To do this, add these lines to
/root/.bash_profile:
PS1="[\u@\h \w]$ "
export PS1
•
If the server slows down, use vmstat, ps (ps axfw), dmesg, top (vztop) to find out what is
happening, never reboot the machine without investigation. If no thinking helps restore the
normal operation, use the Alt+SysRq sequences to dump the memory (showMem) and
processes (showPc).
•
If the server was incorrectly brought down, on its next startup all the partitions will be checked
and quota recalculated for each Container, which dramatically increases the startup time.
•
Do not run any binary or script that belongs to a Container directly from the server, for example,
do not ever do that:
cd /vz/root/99/etc/init.d
./httpd status
Any script inside a Container could have been changed to whatever the Container owner chooses:
it could have been trojaned, replaced to something like rm -rf, etc. You can use only pctl
exec/pctl enter to execute programs inside a Container.
•
Do not use init scripts on the server. An init script may use killall to stop a service, which
means that all similar processes will be killed in all Containers. You can check
/var/run/Service.pid and kill the correspondent process explicitly.
•
You must be able to detect any rootkit inside a Container. It is recommended to use the
chkrootkit package for detection (you can download the latest version from
www.chkrootkit.org), or at least run
rpm -Va|grep "S.5"
to check up if the MD5 sum has changed for any RPM file.
You can also run nmap, for example:
# nmap -p 1-65535 192.168.0.1
Starting nmap V. 2.54BETA22 ( www.insecure.org/nmap/ )
Interesting ports on (192.168.0.1):
245
Troubleshooting
(The 65531 ports scanned but not shown below are in
state: closed)
Port
State
Service
21/tcp
open
ftp
22/tcp
open
ssh
80/tcp
open
http
111/tcp
open
sunrpc
Nmap run completed -- 1 IP address (1 host up) scanned
in 169 seconds
to check if any ports are open that should normally be closed.
That could however be a problem to remove a rootkit from a Container and make sure it is 100%
removed. If you're not sure, create a new Container for that customer and migrate his/her sites and
mail there.
•
Check the /var/log/ directory on the server to find out what is happening on the system.
There are a number of log files that are maintained by the system and Parallels Server Bare
Metal (the boot.log, messages, etc.), but other services and programs may also put their
own log files here depending on your distribution of Linux and the services and applications that
you are running. For example, there may be logs associated with running a mail server (the
maillog file), automatic tasks (the cron file), and others. However, the first place to look into
when you are troubleshooting is the /var/log/messages log file. It contains the boot
messages when the system came up as well as other status messages as the system runs.
Errors with I/O, networking, and other general system errors are reported in this file. So, we
recommend that you read to the messages log file first and then proceed with the other files
from the /var/log/ directory.
•
Subscribe to bug tracking lists. You should keep track of new public DoS tools or remote
exploits for the software and install them into Containers or at servers.
•
When using iptables, there is a simple rule for Chains usage to help protect both the server
and its Containers:
• use INPUT, OUTPUT to filter packets that come in/out the server
• use FORWARD to filter packets that are designated for Containers
246
Troubleshooting
Kernel Troubleshooting
Using ALT+SYSRQ Keyboard Sequences
Press ALT+SYSRQ+H (3 keys simultaneously) and check what is printed at the server console, for
example:
SysRq: unRaw Boot Sync Unmount showPc showTasks showMem loglevel0-8 tErm kIll killalL
Calls Oops
This output shows you what ALT+SYSRQ sequences you may use for performing this or that
command. The capital letters in the command names identify the sequence. Thus, if there are any
troubles with the machine and you're about to reboot it, please press the following sequences
before pressing the Power button:
ALT+SYSRQ+M to dump memory info
ALT+SYSRQ+P to dump processes states
ALT+SYSRQ+S to sync disks
ALT+SYSRQ+U to unmount filesystems
ALT+SYSRQ+L to kill all processes
ALT+SYSRQ+U try to unmount once again
ALT+SYSRQ+B to reboot
If the server is not rebooted after that, you can press the Power button.
247
Troubleshooting
Saving Kernel Faults (OOPS)
You can use the following command to check for the kernel messages that should be reported to
Parallels Server Bare Metal developers:
grep -E "Call Trace|Code" /var/log/messages*
Then, you should find kernel-related lines in the corresponding log file and figure out what kernel
was booted when the oops occurred. Search backward for the "Linux" string, look for strings like
that:
Sep 26 11:41:12 kernel: Linux version 2.6.18-8.1.1.el5.028stab043.1 (root@rhel5-32build) (gcc version 4.1.1 20061011 (Red Hat 4.1.1-30)) #1 SMP Wed Aug 29 11:51:58 MSK
2007
An oops usually starts with some description of what happened and ends with the Code string.
Here is an example:
Aug 25 08:27:46 boar BUG: unable to handle kernel NULL pointer dereference at virtual
address 00000038
Aug 25 08:27:46 boar printing eip:
Aug 25 08:27:46 boar f0ce6507
Aug 25 08:27:46 boar *pde = 00003001
Aug 25 08:27:46 boar Oops: 0000 [#1]
Aug 25 08:27:46 boar SMP
Aug 25 08:27:46 boar last sysfs file:
Aug 25 08:27:46 boar Modules linked in: snapapi26(U) bridge(U) ip_vzredir(U) vzredir(U)
vzcompat(U) vzrst(U) i
p_nat(U) vzcpt(U) ip_conntrack(U) nfnetlink(U) vzfs(U) vzlinkdev(U) vzethdev(U)
vzevent(U) vzlist(U) vznet(U) vzstat(U) vzmo
n(U) xt_tcpudp(U) ip_vznetstat(U) vznetstat(U) iptable_mangle(U) iptable_filter(U)
ip_tables(U) vztable(U) vzdquota(U) vzdev(U) autofs4(U) hidp(U) rfcomm(U) l2cap(U)
bluetooth(U) sunrpc(U) ipv6(U) xt_length(U) ipt_ttl(U) xt_tcpmss(U) ipt_TCPMSS(U)
xt_multiport(U) xt_limit(U) ipt_tos(U) ipt_REJECT(U) x_tables(U) video(U) sbs(U)
i2c_ec(U) button(U) battery(U) asus_acpi(U) ac(U) lp(U) floppy(U) sg(U) pcspkr(U)
i2c_piix4(U) e100(U) parport_pc(U) i2c_core(U) parport(U) cpqphp(U) eepro100(U) mii(U)
serio_raw(U) ide_cd(U) cdrom(U) ahci(U) libata(U) dm_snapshot
(U) dm_zero(U) dm_mirror(U) dm_mod(U) megaraid(U) sym53c8xx(U) scsi_transport_spi(U)
sd_mod(U) scsi_mod(U) ext3(U) jbd(U) ehci_hcd(U) ohci_hcd(U) uhci_hcd(U)
Aug 25 08:27:46 boar CPU: 1, VCPU: -1.1
Aug 25 08:27:46 boar EIP: 0060:[<f0ce6507>] Tainted: P VLI
Aug 25 08:27:46 boar EFLAGS: 00010246 (2.6.18-028stab043.1-ent #1)
Aug 25 08:27:46 boar EIP is at clone_endio+0x29/0xc6 [dm_mod]
Aug 25 08:27:46 boar eax: 00000010
ebx: 00000001
ecx: 00000000
edx: 00000000
Aug 25 08:27:46 boar esi: 00000000
edi: b6f52920
ebp: c1a8dbc0
esp: 0b483e38
Aug 25 08:27:46 boar ds: 007b
es: 007b
ss: 0068
Aug 25 08:27:46 boar Process swapper (pid: 0, veid: 0, ti=0b482000 task=05e3f2b0
task.ti=0b482000)
Aug 25 08:27:46 boar Stack: 0b52caa0 00000001 00000000 b6f52920 00000000f0ce64de
00000000 02478825
Aug 25 08:27:46 boar 00000000 c18a8620 b6f52920 271e1a8c 024ca03800000000 00000000
00000000
Aug 25 08:27:46 boar 00000000 00000000 c18a3c00 00000202 c189e89400000006 00000000
05cb7200
Aug 25 08:27:46 boar Call Trace:
Aug 25 08:27:46 boar [<f0ce64de>] clone_endio+0x0/0xc6 [dm_mod]
Aug 25 08:27:46 boar [<02478825>] bio_endio+0x50/0x55
Aug 25 08:27:46 boar [<024ca038>] __end_that_request_first+0x185/0x47c
Aug 25 08:27:46 boar [<f0c711eb>] scsi_end_request+0x1a/0xa9 [scsi_mod]
248
Troubleshooting
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
Aug 25 08:27:46 boar
3c 83 7f 20 00 8b 45
24 08 0f 85 86 00 00
Aug 25 08:27:46 boar
SS:ESP0068:0b483e38
Aug 25 08:27:46 boar
[<02458f04>] mempool_free+0x5f/0x63
[<f0c713c3>] scsi_io_completion+0x149/0x2f3 [scsi_mod]
[<f0c333b9>] sd_rw_intr+0x1f1/0x21b [sd_mod]
[<f0c6d3b9>] scsi_finish_command+0x73/0x77 [scsi_mod]
[<024cbfa2>] blk_done_softirq+0x4d/0x58
[<02426452>] __do_softirq+0x84/0x109
[<0242650d>] do_softirq+0x36/0x3a
[<024050b7>] do_IRQ+0xad/0xb6
[<024023fa>] default_idle+0x0/0x59
[<0240242b>] default_idle+0x31/0x59
[<024024b1>] cpu_idle+0x5e/0x74
=======================
Code: 5d c3 55 57 89 c7 56 89 ce 53 bb 01 00 00 00 83 ec 0c 8b 68
00 8b 00 89 44 24 04 8b 45 04 89 04 24 8b 40 04 <8b> 40 28 89 44
00 f6 47 10 01 75 0a 85 c9
EIP: [<f0ce6507>] clone_endio+0x29/0xc6 [dm_mod]
Kernel panic - not syncing: Fatal exception in interrupt
All you need is to put the oops into a file and then send this file as part of your problem report to
the Parallels support team.
Finding a Kernel Function That Caused the D Process State
If there are too many processes in the D state and you can't find out what is happening, issue the
following command:
# objdump -Dr /boot/vmlinux-`uname -r` >/tmp/kernel.dump
and then get the process list:
# ps axfwln
F UID
PID PPID PRI NI VSZ RSS WCHAN STAT
100
0 20418 20417 17 0 2588 684
- R
100
0
1
0
8 0 1388 524 145186 S
040
0 8670
1
9 0 1448 960 145186 S
040
0 8713
1 10 0 1616 1140 11ea02 S
TTY
?
?
?
?
TIME
0:00
0:00
0:00
0:00
COMMAND
ps axfwln
init
syslogd -m 0
crond
Look for a number under the WCHAN column for the process in question. Then, open
/tmp/kernel.dump in an editor, find that number in the first column and then scroll backward to
the first function name, which can look like this:
"c011e910 <sys_nanosleep>:"
Then you can tell if the process “lives” or is blocked into the found function.
Problems With Container Management
This section includes recommendations on how to settle some problems with Containers.
249
Troubleshooting
Failure to Start a Container
An attempt to start a Container fails.
Solution 1
If there is a message on the system console: parameters missing, and the list of missed
parameters follows the message, set these parameters using the pctl set --save command
(see Performing Initial Configuration (p. 34) for instructions). Try to start the Container once
again.
Solution 2
If there is a message on the system console: IP address is already used, issue the cat
/proc/vz/veinfo command. The information about the Container numeric identifier, Container
class, number of Container’s processes and Container IP address shall be displayed for each
running Container. This shall also demonstrate that your Container is up, i.e. it must be running
without any IP address assigned. Set its IP address using the command:
pctl set CT_ID --ipadd IP_addr --save
where CT_ID represents the Container numeric identifier and IP_addr represents an actual IP
address.
Solution 3
Poor UBC parameters might prevent the Container from starting. Try to validate the Container
configuration (see Validating Container Configuration (p. 152)). See what configuration
parameters have caused the error and set appropriate values using the pctl set --save
command.
Solution 4
The Container might have used all its disk quota (either disk space or disk inodes). Check the
Container disk quota (see the Managing Disk Quotas section (p. 115) and Chapter 4 for details)
and increase the quota parameters if needed (see Setting Up Per-Container Disk Quota
Parameters (p. 118)).
Solution 5
Run the vzfsutil utility to make sure that the VZFS symlinks inside the Container work correctly.
For example:
vzfsutil --call –t /vz/template /vz/private/<CT_ID>
The complete reference on the vzfsutil utility is provided in the Parallels Command Line
Reference Guide.
Solution 6
250
Troubleshooting
The Container administrator might have inadvertently modified, replaced, or deleted any file that is
part of an application or OS template, which has brought about the Container malfunction. In this
case, restore the file(s) with the pctl recover command (see the Reinstalling Container
section (p. 76) for details).
Solution 7
Restore the latest operable copy of the Container by means of the vzrestore utility (see the
Managing virtual machine and Container Backups section (p. 44) for details).
Failure to Access a Container From Network
Solution 1
The IP address assigned to the Container might be already in use in your network. Make sure it is
not. The problem Container address can be checked by issuing the following command:
# grep IP_ADDRESS /etc/vz/conf/<CT_ID>.conf
IP_ADDRESS="10.0.186.101"
The IP addresses of other Containers, which are running, can be checked by running
cat /proc/vz/veinfo
Solution 2
Make sure the routing to the Container is properly configured. Containers can use the default router
for your network, or you may configure the server as rooter for its Containers.
Failure to Log In to a Container
The Container starts successfully, but you cannot log in.
Solution 1
You are trying to connect via SSH, but access is denied. Probably you have not set the password
of the root user yet or there is no such user. In this case, use the pctl set --userpasswd
command. For example, for Container 101 you might issue the following command:
# pctl set 101 --userpasswd root:secret
Solution 2
Check forwarding settings by issuing the following command:
# cat /proc/sys/ipv4/conf/venet0/forwarding
If it is 0 then change it to 1 by issuing the following command:
# echo 1 > /proc/sys/ipv4/conf/venet0/forwarding
251
Troubleshooting
Getting Technical Support
This section provides information on how to get technical support from Parallels.
Preparing and Sending Questions to Technical Support
In most cases, the support team must rely on the customer's observations and communications
with the customer to diagnose and solve the problem. Therefore, the detailed problem report is
extremely important. You can submit a support report by visiting the
http://www.parallels.com/en/support/virtuozzo/request/ web page and filling in the Online Support
Form. When describing the problem, please do mention the following:
•
symptoms of the problem
•
when the problem began including the circumstances of the failure
•
any changes you made to your system
•
other information that may be relevant to your situation (e.g. the installation method)
•
specific hardware devices that may be relevant to your problem
You can also make use of the Parallels Helpdesk support tool. To do this:
1
Follow the https://helpdesk.parallels.com/ link.
2
Register with the Parallels Helpdesk (if you have not done so before) by clicking the Get
Access to Parallels RT link on the Helpdesk login page and following the instructions provided
on the Activate Your Support Account screen.
3
Log in to the Helpdesk using the received credentials.
4
At the top of the RT At Glance screen, select the Parallels Server Bare Metal component your
problem relates to on the drop-down menu, and click the New Ticket in button:
5
On the Create New Ticket screen, fill in the appropriate fields, describe your problem, and
click the Create button to make a new support ticket.
Another way of getting help is to directly call us or visit one of our offices. The information about
phone numbers, contact people and office addresses is available on the contact pages at
http://www.parallels.com/en/contact and http://www.parallels.com/en/support/phone/.
252
Troubleshooting
Submitting Problem Report to Technical Support
Parallels Server Bare Metal is shipped with a special utility, vzreport, allowing you to compile a
detailed report if you have any problems and to automatically send it to the Parallels support team.
After receiving your report, the support team will closely examine your problem and make its best
to solve it as quickly as possible.
vzreport has two modes of execution—full screen and command line. By default, the utility starts
in the full screen mode. However, you can force the utility to run in the command line mode by
specifying any option containing your contact information (e.g., -n denoting your name) or the
problem report description (e.g., -m used to provide additional information on your problem).
Detailed information on all the options that can be passed to vzreport in the command line is
provided in the Parallels Command Line Reference Guide.
After running the vzreport utility in the full screen mode, the Problem Report Wizard is opened,
which will guide you through a number of steps asking you to provide the necessary information to
generate a problem report. On the Welcome screen, just click Next to proceed with the wizard.
You will be presented with the following window:
In this window, you should enter your name, e-mail, and the name of your company into the
corresponding fields. Make sure that you type a valid e-mail address. Otherwise, the Parallels
support team will not be able to contact you. In the Subject field, specify what problem you
encountered. You can also provide additional information in the Problem description field which,
in your opinion, can help solve the problem.
253
Troubleshooting
Clicking Next in the Your contact information and issue description window starts collecting
Parallels Server Bare Metal logs and the information on your system and network settings into a
special file. You can view the progress in the Gathering Information window. This file will be sent
to the Parallels support team upon the completion of the wizard. The file does not contain any
private information!
After the utility has gathered all the necessary information on your server, the Submit report
window is displayed:
In this window, you can do one of the following:
•
Click the Submit button to send your problem report to the Parallels technical support team.
The report is dispatched directly to Parallels by using the HTTP protocol and port 80. However,
if you use an HTTP proxy server for handling all your HTTP requests and wish your problem
report to be sent via this server, you should specify the hostname or IP address of the server in
the /etc/vz/vz.conf configuration file on the server as the value of the HTTP_PROXY
parameter. After the problem report has been successfully sent to the Parallels support, the
Congratulations window is displayed informing you:
• Of the ID assigned to your report. Use this ID every time you communicate with the Parallels
support via e-mail or the Parallels Helpdesk support tool
• That an e-mail message providing you with detailed information on your problem report has
been sent to the e-mail address you specified in the E-mail field of the Your contact
information and issue description window.
Click the Cancel button if you do not wish to dispatch the problem report to the support team at
the moment. You can do it later on by manually sending the generated zip file to the Parallels
support team. The full path to this file is indicated in the Submit report window.
Note: Your physical server must be assigned at least one public IPv4 address for the vzreport utility to
automatically send the generated report to the Parallels support team.
254
Troubleshooting
Establishing Secure Channel to Parallels Support
Parallels Server Bare Metal provides you with a special tool - Support Tunnel - which allows you to
establish a private secure channel to the Parallels support team server. After establishing such a
channel, the support team will be able to quickly and securely connect to your Parallels server and
diagnose and solve your problem. The secure connection to your server is achieved through a
Virtual Private Network (VPN) created between the Parallels support team server and your server.
To start using the Virtuozzo Support Tunnel tool:
•
Make sure the openvpn (version 2.0 and above) and vzvpn packages are installed on your
server. These packages are automatically installed during the Parallels Server Bare Metal
installation.
•
Make sure that port 80 is opened on the server.
•
Edit the /etc/vzvpn/vzvpn.conf file to specify the correct parameters for your proxy
server, if you use any. Detailed information on these parameters is given in the vzvpn
Configuration File subsection of the Parallels Command Line Reference Guide.
After you have completed the tasks above and in case you encountered a problem, you can do the
following to get assistance from the Parallels support:
1
Obtain a special certificate from Parallels which will uniquely identify you as a Parallels Server
Bare Metal user. Certificates are issued by Parallels in the form of files and should be installed
on your server by issuing the vzvpn.sh key-install certificate command where
certificate denotes the name of the certificate file obtained from Parallels. You can get a
certificate in one of the following ways:
• Visit the http://www.parallels.com/en/support/virtuozzo/certificates web site, fill up the
Request Secure Virtuozzo Support Tunnel Certificate form, and click the Submit button.
After a while, a certificate will be generated and sent to the email address you provided in
the Request Secure Virtuozzo Support Tunnel Certificate form.
• Contact the Parallels support team via e-mail or by telephone and ask for a valid certificate.
2
After you are ready with the certificate installation, make sure your server is connected to the
Internet.
3
On the server, execute the /etc/init.d/vzvpn.sh start command to establish a VPN
between your server and the Parallels support server.
4
Contact the Parallels support team (by telephone or via e-mail) and inform them of the problem
you encountered. You should also mention that you have launched the Virtuozzo Support
Tunnel tool and established a VPN to the Parallels support server.
5
After that, the Parallels support team will connect to your server by using the secure VPN
established, closely examine your problem, and make its best to solve the problem as quickly
as possible.
255
Troubleshooting
Notes:
1. Support Tunnel is implemented as a standard Linux service running in the background of your
system. Therefore, to have this service running after your server reboot, you should set it to the
autoboot mode or start it manually again by executing the /etc/init.d/vzvpn start command.
2. To close the VPN session with the Parallels support server, you should issue the
/etc/init.d/vzvpn stop command on the server.
256
CHAPTER 11
Glossary
This glossary defines terms and spells out abbreviations used in Parallels Server Bare Metal
documentation. References to terms defined elsewhere in the glossary appear in italics.
Application template. A template used to install a set of applications in Containers. See also
Template.
Container (or regular Container). A virtual private server, which is functionally identical to an
isolated standalone server, with its own IP addresses, processes, files, its own users database, its
own configuration files, its own applications, system libraries, and so on. Containers share one
Parallels server and one OS kernel. However, they are isolated from each other. A Container is a
kind of ‘sandbox’ for processes and users.
Guest operating system (Guest OS). An operating system installed inside a virtual machine and
Container. It can be any of the supported Windows or Linux operating systems.
Hardware virtualization. A virtualization technology allowing you to virtualize physical servers at
the hardware level. Hardware virtualization provides the necessary environment for creating and
managing Parallels virtual machines.
Operating system virtualization (OS virtualization). A virtualization technology allowing you to
virtualize physical servers at the operating system (kernel) level. OS virtualization provides the
necessary environment for creating and managing Parallels Containers.
OS template (Operating System template). A template used to create new Containers with a
pre-installed operating system. See also Template.
Package set. See Template.
Parallels Management Console. A Parallels Server Bare Metal management and monitoring tool
with graphical user interface. Parallels Management Console is cross–platform and can run on
Microsoft Windows and Linux computers.
Parallels Server. A hardware virtualization solution that enables you to efficiently use your physical
server's hardware resources by sharing them between multiple virtual machines created on this
server.
Parallels server (physical server or server). A server where the Parallels Server Bare Metal
software is installed for hosting Parallels virtual machines and Containers. Sometimes, it is marked
as Container 0.
Glossary
Parallels Server Bare Metal license. A special license that you should install on the physical
server to be able to start using Parallels Server Bare Metal. Every physical server must have its own
license installed.
Parallels Virtuozzo Containers for Linux. An operating system virtualization solution allowing you
to create multiple isolated Containers on a single physical server to share hardware, licenses, and
management effort with maximum efficiency.
Private area. A part of the file system storing Container files that are not shared with other
Containers.
Template (package set). A set of original application files (packages) repackaged for mounting
over Virtuozzo File System. There are two types of templates. OS Templates are used to create
new Containers with a pre-installed operating system. Application templates are used to install an
application or a set of applications in Containers.
UBC. An abbreviation of User Beancounter.
User Beancounter. The subsystem of the Parallels Server Bare Metal software for managing
Container memory and some system-related resources.
Virtual Environment (VE). An obsolete designation of a Container.
Virtuozzo File System (VZFS). A virtual file system for mounting to Container private areas. VZFS
symlinks are seen as real files inside Containers.
Virtual machine (VM). A computer emulated by Parallels Server Bare Metal. Like a Container, a
virtual machine is functionally identical to an isolated standalone computer, with its own IP
addresses, processes, files, its own users database, its own configuration files, its own
applications, system libraries, and so on. However, as distinct from Containers, virtual machines
run their own operating systems rather than sharing one operating system kernel.
258
Index
Index
A
About This Guide - 9
Adding a New Device - 94
Advanced Tasks - 213
Aligning Disks and Partitions in Virtual
Machines - 236
Applying New Configuration Sample to
Container - 153
Assigning USB Devices to Virtual Machines 102
Associating Container Files With Application
Templates - 124
Available Capabilities for Containers - 215
B
Backups Overview - 45
Basics of Hardware Virtualization - 20
Basics of OS Virtualization - 17
C
Capabilities Defined by POSIX Draft - 216
Changing Services Mode - 161
Changing System Time From Containers 225
Changing the Disk Type - 88
Checking Quota Status - 121
Choosing a Container ID - 33
Choosing OS EZ Template - 34
Cleaning Up Containers - 122
Compacting Disks - 91
Compatibility With Parallels Products - 32
Configuring Additional Memory Parameters 145
Configuring Capabilities - 213
Configuring CPU Affinity for Virtual Machines
and Containers - 115
Configuring IP Address Ranges for Host-Only
Networks - 105
Configuring Legacy Containers - 143
Configuring Main Memory Resources - 144
Configuring Main VSwap Parameters - 140
Configuring Network Bandwidth Management
- 133
Configuring Network Classes - 127
Configuring Network Settings - 35
Configuring Passwordless Access to the
Source Node - 53
Configuring Priority Levels for Virtual
Machines and Containers - 135
Configuring the Disk I/O Bandwidth - 136
Configuring the Memory Allocation Limit - 141
Configuring the Number of CPUs - 113
Configuring veth Adapter Parameters - 181
Configuring Virtual Adapter Parameters - 185
Configuring Virtual Devices - 98
Configuring Virtual Machines and Containers
in Host-Routed Mode - 173
Configuring Virtual Machines and Containers
in Virtual Network Mode - 174
Configuring Virtual Network Parameters - 176
Connecting Containers to Virtual Networks 182
Connecting Virtual Machines to Virtual
Networks - 186
Connecting Virtual Networks to Adapters 178
Converting Third-Party Virtual Machines and
Disks - 106
Copying a Virtual Machine and Container
Within the Server - 40
Create a Template - 86
Creating a Snapshot - 83
Creating a Virtual Machine and Container - 29
Creating a Virtual Network - 175
Creating and Deleting veth Network Adapters
- 180
Creating and Deleting Virtual Adapters - 184
Creating Configuration Files for New Linux
Distributions - 234
Creating Customized Containers - 218
Creating VLAN Adapters - 166
Index
Creating VZFS Symlinks Inside a Container 214
Customizing Container Reinstallation - 78
Customizing the /proc/meminfo Output in
Containers - 229
D
Deleting a Device - 100
Deleting a Snapshot - 85
Deleting a Virtual Machine and Container - 44
Deleting Virtual Networks - 179
Deploying a Template - 87
Detaching Containers From Caches - 125
Determining Container Identifiers by Process
IDs - 162
Differences Between Host-Routed and Virtual
Network Modes - 172
Disabling Container - 75
Disk Quota Parameters - 116
Documentation Conventions - 11
E
Enabling CPU Hotplug for Virtual Machines 114
Enabling Memory Hotplug for Virtual
Machines - 148
Enabling VNC Access to Virtual Machines 108
Enabling VPN for Containers - 226
Establishing Secure Channel to Parallels
Support - 255
F
Failure to Access a Container From Network 251
Failure to Log In to a Container - 251
Failure to Start a Container - 250
Feedback - 13
Finding a Kernel Function That Caused the D
Process State - 249
G
General Considerations - 245
General Migration Requirements - 55
Getting Help - 12
Getting Technical Support - 252
Glossary - 257
H
Hardware Virtualization Layer - 19
Host-Routed Mode - 167
I
Increasing the Disk Capacity - 89
Initializing the Newly Added Disk - 96
Installing the License - 194
Introduction - 9
K
Keeping Your System Up To Date - 201
Kernel Troubleshooting - 247
L
Learning Parallels Server Bare Metal 5.0
Basics - 14
Learning Private Networks - 187
License Statuses - 200
Linux-Specific Capabilities - 217
Listing Adapters - 165
Listing Snapshots - 84
Listing Templates - 87
Listing Virtual Machines and Containers - 38
Listing Virtual Networks - 177
Loading iptables Modules - 231
Loading iptables Modules to Containers - 232
Loading iptables Modules to Parallels Server 231
M
Main Operations on Services and Processes 156
Making Screenshots - 101
Managing Adapters in Containers - 179
Managing Adapters in Virtual Machines - 183
Managing Container Resources Configuration
- 149
Managing CPU Resources - 110
Managing CPU Shares and Limits - 111
Managing Disk I/O Parameters - 134
Managing Disk Interfaces - 92
Managing Disk Quotas - 115
Managing Licenses - 193
Managing Memory Parameters for Containers
- 139
Index
Managing Memory Resources for Virtual
Machines - 144
Managing Network Accounting and
Bandwidth - 125
Managing Network Adapters on the Parallels
Server - 164
Managing Parallels Server Bare Metal
Network - 163
Managing Private Networks - 187
Managing Processes and Services - 157
Managing Resources - 109
Managing Server Resources Parameters 227
Managing Services and Processes - 154
Managing Snapshots - 82
Managing Templates - 86
Managing Virtual Machine and Container
Backups - 44
Managing Virtual Machine Devices - 93
Managing Virtual Machine Disks - 88
Managing Virtual Machines and Containers 28
Managing Virtual Networks - 174
Migrating Containers - 61
Migrating Containers from a Server with
Parallels Virtuozzo Containers - 62
Migrating Containers to Virtual Machines - 61
Migrating Physical Computers to Virtual
Machines and Containers - 64
Migrating Virtual Machines and Containers 54
Migrating Virtual Machines and Containers
Between Parallels Servers - 56
Migrating Virtual Machines to Containers - 70
Migrating Xen Virtual Machines - 71
Migration Restrictions for Containers - 67
Monitoring Processes in Real Time - 160
Monitoring Resources - 235
Mounting Virtual Machines - 104
Moving Container Files to the Cache Area 123
Moving Container Within the Parallels Server 74
N
Network Traffic Parameters - 126
Networking Modes in Parallels Server Bare
Metal - 166
O
Obtaining Server ID From Inside a Container 226
Organization of This Guide - 10
OS Virtualization Layer - 16
P
Parallels Containers - 17
Parallels Management Console - 25
Parallels Server Bare Metal 5.0 Overview - 15
Parallels Server Bare Metal Configuration - 19
Parallels Virtual Machines - 21
Pausing a Virtual Machine - 81
Performing Container-Specific Operations 72
Performing Initial Configuration - 34
Performing Virtual Machine-Specific
Operations - 80
Physical Server Availability Considerations 27
Preparing and Sending Questions to
Technical Support - 252
Problems With Container Management - 249
R
Reducing the Disk Capacity - 90
Reinstalling Container - 76
Requirements for Migrating to Containers 66
Requirements for Migrating to Virtual
Machines - 69
Resource Management - 26
Reverting to a Snapshot - 85
Running Commands in a Virtual Machine and
Container - 43
Running Parallels Server Bare Metal 5.0 in
Virtual Machines - 242
S
Saving Kernel Faults (OOPS) - 248
Scaling Container Configuration - 151
Setting Immutable and Append Flags for
Container Files and Directories - 228
Setting Name for Container - 73
Setting Passwords for Virtual Machines and
Containers - 36
Setting Startup Parameters - 36
Index
Setting Up Per-Container Disk Quota
Parameters - 118
Setting Up Private Networks - 191
Setting Up Second-Level Disk Quota
Parameters - 120
Splitting Server Into Equal Pieces - 150
Standard Migration - 57
Starting, Stopping, and Querying Status of a
Virtual Machine and Container - 37
Starting, Stopping, and Restarting Services 162
Storing Extended Information on a Virtual
Machine and Container - 39
Submitting Problem Report to Technical
Support - 253
Support of Virtual and Real Media - 24
Supported Guest Operating Systems - 31
Suspending a Virtual Machine and Container
- 42
T
Templates - 18
Transferring the License to Another Server 196
Troubleshooting - 244
Tuning VSwap - 142
Turning On and Off Network Bandwidth
Management - 131
Turning On and Off Per-Container Disk
Quotas - 117
Turning On and Off Second-Level Quotas for
a Container - 119
U
Understanding Licensing - 26
Updating Containers - 210
Updating EZ Template Packages Inside a
Container - 211
Updating EZ Templates - 206
Updating in Command-Line Mode - 209
Updating in Graphical Mode - 203
Updating OS EZ Template Caches - 212
Updating Parallels Server Bare Metal
Software - 202
Updating Software In Virtual Machines - 210
Updating System Files - 204
Updating the Current License - 195
Using ALT+SYSRQ Keyboard Sequences 247
Using Customized Application Template 223
Using Customized OS EZ Templates - 219
Using EZ OS Template Sets - 221
Using pbackup and prestore - 50
Using pctl backup and pctl restore - 47
V
Validating Container Configuration - 152
Viewing Active Processes and Services - 158
Viewing Detailed Information About Virtual
Machines - 80
Viewing Disk I/O Statistics - 138
Viewing Network Traffic Statistics - 129
Viewing the Current License - 197
Viewing the License - 198
Virtual Machine Files - 23
Virtual Machine Hardware - 22
Virtual Network Mode - 170
Virtuozzo File System - 18
W
What are Disk Quotas? - 116
What are Resource Control Parameters? 109
What Are Services and Processes - 155
Z
Zero-Downtime Migration - 59
Download PDF
Similar pages