As part of an effort to improve its product lines, EMC periodically releases revisions of its software and hardware. Therefore, some functions that are described in this document might not be supported by all versions of the software or hardware currently in use. The product release notes provide the most up-to-date information on product features.

Contact your EMC technical support professional if a product does not function correctly or does not function as described in this document.

Note

This document was accurate at publication time. Go to EMC Online Support ( https:// support.emc.com

) to ensure that you are using the latest version of this document.

Purpose

This document describes how to configure and use EMC NetWorker.

Audience

This guide is part of the NetWorker documentation set, and is intended for use by system administrators who are responsible for setting up and maintaining backups on a network. Operators who monitor daily backups will also find this guide useful.

Revision history

The following table presents the revision history of this document.

Table 1 Revision history

Revision

Date Description

February 3 , 2017 Third release of this document for EMC NetWorker 9.1.

Includes the following updates: l l l l

Updated the Creating a custom backup script section in the Backing Up the Data chapter to include information about declaring environment variables.

Updated the section "Is the bootstrap on a remote device" in the Recovery chapter.

Updated the section "Processes on NetWorker hosts" in the Overview chapter with more information about the nsrindexd process.

The path to the NMC staging directory for Linux in the "Recovery" chapter.

January 5, 2017

December 22,

2016

Fixed description of VMware Protection options to include NetWorker VMware Protection with the vProxy appliance, or NVP.

First release of this document for EMC NetWorker 9.1.

EMC NetWorker 9.1 Administration Guide 21

Preface

Related documentation

The NetWorker documentation set includes the following publications, available on

EMC Online Support: l l l l l l l l l l l l l l l

EMC NetWorker Online Software Compatibility Matrix

Provides a list of client, server, and storage node operating systems supported by the EMC information protection software versions. You can access the matrix at http://compatibilityguide.emc.com:8080/CompGuideApp/ .

EMC NetWorker Administration Guide

Describes how to configure and maintain the NetWorker software.

EMC NetWorker Network Data Management Protocol (NDMP) User Guide

Describes how to use the NetWorker software to provide data protection for

NDMP filers.

EMC NetWorker Cluster Integration Guide

Contains information related to configuring NetWorker software on cluster servers and clients.

EMC NetWorker Installation Guide

Provides information on how to install, uninstall, and update the NetWorker software for clients, storage nodes, and servers on all supported operating systems.

EMC NetWorker Updating from a Previous Release Guide

Describes how to update the NetWorker software from a previously installed release.

EMC NetWorker Release Notes

Contains information on new features and changes, fixed problems, known limitations, environment and system requirements for the latest NetWorker software release.

EMC NetWorker Command Reference Guide

Provides reference information for NetWorker commands and options.

EMC NetWorker Data Domain Boost Integration Guide

Provides planning and configuration information on the use of Data Domain devices for data deduplication backup and storage in a NetWorker environment.

EMC NetWorker Performance Optimization Planning Guide

Contains basic performance tuning information for NetWorker.

EMC NetWorker Server Disaster Recovery and Availability Best Practices Guide

Describes how to design, plan for, and perform a step-by-step NetWorker disaster recovery.

EMC NetWorker Snapshot Management Integration Guide

Describes the ability to catalog and manage snapshot copies of production data that are created by using mirror technologies on EMC storage arrays.

EMC NetWorker Snapshot Management for NAS Devices Integration Guide

Describes how to catalog and manage snapshot copies of production data that are created by using replication technologies on NAS devices.

EMC NetWorker Security Configuration Guide

Provides an overview of security configuration settings available in NetWorker, secure deployment, and physical security controls needed to ensure the secure operation of the product.

EMC NetWorker VMware Integration Guide

Provides planning and configuration information on the use of VMware in a

NetWorker environment.

EMC NetWorker 9.1 Administration Guide

Preface l l l l l l l

EMC NetWorker Error Message Guide

Provides information on common NetWorker error messages.

EMC NetWorker Licensing Guide

Provides information about licensing NetWorker products and features.

EMC NetWorker REST API Getting Started Guide

Describes how to configure and use the NetWorker REST API to create programmatic interfaces to the NetWorker server.

EMC NetWorker REST API Reference Guide

Provides the NetWorker REST API specification used to create programmatic interfaces to the NetWorker server.

EMC NetWorker 9.1 with EMC CloudBoost 2.1 Integration Guide

Describes the integration of NetWorker with CloudBoost.

EMC NetWorker Management Console Online Help

Describes the day-to-day administration tasks performed in the NetWorker

Management Console and the NetWorker Administration window. To view the online help, click Help in the main menu.

EMC NetWorker User Online Help

Describes how to use the NetWorker User program, which is the Windows client interface, to connect to a NetWorker server to back up, recover, archive, and retrieve files over a network.

Special notice conventions that are used in this document

EMC uses the following conventions for special notices:

NOTICE

Identifies content that warns of potential business or data loss.

Note

Contains information that is incidental, but not essential, to the topic.

Typographical conventions

EMC uses the following type style conventions in this document:

Table 2 Style conventions

Bold

Italic

Monospace

Monospace italic

Monospace bold

[ ]

Used for names of interface elements, such as names of buttons, fields, tab names, and menu paths (what the user specifically selects or clicks)

Used for full titles of publications that are referenced in text

Used for: l l l l

System code

System output, such as an error message or script

Pathnames, file names, prompts, and syntax

Commands and options

Used for variables

Used for user input

Square brackets enclose optional values

EMC NetWorker 9.1 Administration Guide 23

Preface

Table 2 Style conventions (continued)

{ }

...

Vertical bar indicates alternate selections - the bar means “or”

Braces enclose content that the user must specify, such as x or y or z

Ellipses indicate non-essential information that is omitted from the example

Where to get help

EMC support, product, and licensing information can be obtained as follows:

Product information

For documentation, release notes, software updates, or information about EMC products, go to EMC Online Support at https://support.emc.com

Technical support

Go to EMC Online Support and click Service Center. Several options for contacting

EMC Technical Support appear on the site. Note that to open a service request, you must have a valid support agreement. Contact your EMC sales representative for details about obtaining a valid support agreement or with questions about your account.

Online communities

Go to the EMC Community Network at https://community.emc.com

for peer contacts, conversations, and content on product support and solutions. Interactively engage online with customers, partners, and certified professionals for all EMC products.

Your comments

Your suggestions help to improve the accuracy, organization, and overall quality of the user publications. Send your opinions of this document to

[email protected]

24 EMC NetWorker 9.1 Administration Guide

CHAPTER 1 Overview

This chapter contains the following topics: l l l

The NetWorker environment

..............................................................................26

NetWorker services

........................................................................................... 28

NetWorker user interfaces

.................................................................................36

Overview 25

Overview

The NetWorker environment

The EMC

NetWorker

environment provides the ability to protect an enterprise against data loss. As the enterprise grows, so does the complexity and importance of protecting data. The NetWorker software provides the power and flexibility to meet these challenges.

The NetWorker software is a cross-platform, client/server application that provides the ability to remotely manage all NetWorker servers from a web-enabled, graphical interface.

NetWorker components

Several components make up the NetWorker environment and provide the ability to protect against data loss.

The following figure illustrates the main components in a NetWorker environment.

Figure 1 NetWorker components

NMC Server

The NetWorker Management Console (NMC) server or Console server is a Java-based web application and database server. The NMC Server manages all NetWorker

Servers and Clients. The NMC Server also provides reporting and monitoring capabilities for all NetWorker Servers and Clients in the environment. NMC Server relies on the NetWorker Authentication Service for user account authentication.

26 EMC NetWorker 9.1 Administration Guide

Overview

Datazone

A NetWorker datazone is composed of a single NetWorker Server, its clients, and storage nodes. You can add additional datazones as backup requirements increase.

NetWorker Authentication Service

The NetWorker Authentication Service provides centralized token-based authentication to components in a NetWorker 9.1.x environment. You can configure the NetWorker Authentication Service to use a local user database or external identity providers (LDAP and AD) for authentication.

NetWorker Server

The NetWorker Server is a collection of processes and programs that are installed on a host that performs NetWorker services. The NetWorker Server also acts as a storage node and can control multiple remote storage nodes.

NetWorker client

A NetWorker client is a physical computer that you install the NetWorker client software on. The NetWorker client computer can be any computer in a datazone that contains data you want to back up. The NMC server, NetWorker server, and

NetWorker storage node hosts are also NetWorker clients.

NetWorker client resource overview

A NetWorker client resource defines the data that you want to back up on a host. You can create multiple client resources for a NetWorker host, and each resource defines a different dataset.

EMC provides the NetWorker client software for a variety of operating system platforms. Any NetWorker server can backup a NetWorker client, regardless of the platform the client resides on. For example, you can back up a NetWorker client on a

Microsoft Windows computer to a NetWorker server on a Solaris computer.

NetWorker Storage Node

NetWorker can back up data to local devices on a NetWorker Server or remote devices on a storage node. A storage node controls storage devices such as tape drives, disk devices, autochangers, and silos.

The NetWorker Server is a local storage node. Use a remote storage node to offload most of the data movement in a backup or a recovery operation from the NetWorker

Server. A remote storage node improves performance, but it requires high I/O bandwidth to manage data transfer from local clients or network clients to target devices. The operating system of a remote storage node can differ from the

NetWorker Server.

NetWorker REST API

The NetWorker REST API is an interface that allows customer to access the

NetWorker data protection service and to build client applications that automate

NetWorker operations. The EMC NetWorker REST API Getting Started Guide describes how to use NetWorker REST API, and the EMC NetWorker REST API Reference Guide provides a full description of the API resources.

NetWorker components 27

Overview

EMC Licensing Solution

NetWorker 9.0.x and later servers use the EMC Licensing Solution, an EMC standard for licensing in software products.

The EMC Licensing Solution uses an EMC License Server and reads a license file to determine which products are licensed and how much storage space to request for each datazone in the environment.

All new installations of NetWorker use the EMC Licensing Solution. The chapter "EMC

Licensing Solution" in the EMC NetWorker Licensing Guide provides information on how to implement the EMC Licensing Solution for new and upgraded installations of the

NetWorker software. The "EMC Licensing Solution" chapter also describes the EMC

License Server and the use of the license file.

Restricted datazones

Restricted datazones provide NetWorker administrators with the ability to organize a

NetWorker environment into a multi-tenancy configuration.

In a multi-tenancy configuration, each restricted datazone contains one NetWorker server and other associated NetWorker resources. Global administrators oversee the setup and management of several restricted data zones and assign tenant administrators with access to a restricted datazone. A tenant administrator can only manage NetWorker resources within an assigned restricted datazone.

Deduplication storage systems

The NetWorker software supports backup data deduplication on Data Domain

® storage systems.

The

EMC NetWorker Data Domain Boost Integration Guide provides detailed information about setting up DD Boost deduplication devices to work with NetWorker.

Virtual environments

NetWorker clients can be created for virtual machines for either traditional backup or

VMware Consolidated Backup (VCB). Additionally, the NetWorker software can automatically discover virtual environments and changes to those environments on either a scheduled or on-demand basis and provides a graphical view of those environments.

NetWorker services

The main services and programs for the NetWorker server are the NetWorker storage node, NetWorker client, and NetWorker Management Console (NMC) server.

This section includes information on the NetWorker services, and how to start and stop the services.

For more information about: l l

Main NetWorker services—The EMC NetWorker Command Reference Guide or the

UNIX man pages provides more information.

Service port requirements when configuring a firewall—The EMC NetWorker

Security Configuration Guide provides more information.

28 EMC NetWorker 9.1 Administration Guide

Overview

Processes on NetWorker hosts

Each NetWorker host requires processes to provide configuration and management support of the NetWorker software.

NetWorker authentication service

To support the NetWorker authentication service feature, one or more tomcat processes start on the NetWorker server. The tomcat process provides the authentication service with a database server instance, which enables the authentication service to manage tokens and supports user database management.

NetWorker REST API

The NetWorker REST API service is deployed in the same Apache Tomcat container as NetWorker authentication service. The NetWorker REST API uses the same set of

Tomcat processes to deliver its service.

NetWorker client

The nsrexecd process runs on a NetWorker client. This process authenticates and manages NetWorker server remote execution requests and starts the save and savefs processes on the client to support backup requests.

NetWorker server

The following table summarizes the processes that support the NetWorker server software.

Table 3 NetWorker server processes

Process

nsrd nsrmmdbd nsrjobd nsrindexd

Function

l l l l

NetWorker save and recovery daemon.

The master service that controls other services on the NetWorker server, clients, and storage nodes.

Monitors active save or recover program sessions.

In response to a recover session, nsrd spawns an agent process, ansrd.

l l

NetWorker save and recover media management database service daemon.

Provides media database management services to the local nsrd and nsrmmd services and records entries in the media database.

Monitors NetWorker activity during a backup or recovery operation.

Provides an indexing service to read, write, and remove index entries.

The nsrd service starts one nsrindexd process on the NetWorker server. The nsrindexd process spawns an additional

Processes on NetWorker hosts 29

Overview

Table 3 NetWorker server processes (continued)

Process Function

helper nsrindexd process for each index session.

NetWorker uses index sessions to read, write, or delete index entries, for example, when

NetWorker saves an index, or when a user performs a file-level or browsable recover.

When the read or write operation completes, the helper nsrindexd process closes.

nsrmmgd nsrlogd nsrcpd nsrdispd nsrdisp_nwbg nsrlmc nsrctld l l l

Manages tape library operations.

Provides an RPC-based service that manages all jukebox operations on behalf of the nsrd service.

The nsrd service starts only one instance of nsrmmgd on the NetWorker server as needed.

Supports the NetWorker audit log service, which is configured to run on the NetWorker server by default.

l l

Starts automatically when a user accesses the

Hosts Task

window in the

NetWorker Administration interface.

Allows users to distribute and upgrade

NetWorker and module software from a centralized software repository across a network.

Handles RPC-based calls for the nsrd process, from remote third party processes.

Started by nsrdispd to handle NMC server requests for information from the RAP and media databases on the NetWorker server.

l l l

Supports licensing requests.

For the traditional licensing model, nsrlmc requests a license from the lgtolmd process.

For the CLP/ELMS licensing model, nsrlmc requests capacity and update licenses from the ELMS server.

The top-level NetWorker server process that monitors, stops, and starts all NetWorker server processes.

30 EMC NetWorker 9.1 Administration Guide

Overview

Table 3 NetWorker server processes (continued)

Process

nsrvmwsd tomcat7 (Windows), tomcat (UNIX)

Function

Provides a web service to manage VMware

VM backups that are part of the NetWorker

VMware protection feature.

Tomcat web server instance for the

NetWorker Authentication Service.

NetWorker storage node

The following table summarizes the services that support the NetWorker storage node software.

Table 4 NetWorker storage node processes

Process

nsrmmd nsrsnmd nsrlcpd

Function

l l

Provides device support, generates mount requests, multiplexes save set data during a multi client backup, and demultiplexes recover data. It writes the data sent by save to storage media.

Forwards storage information to the nsrmmdbd process on the NetWorker server, which the NetWorker server adds to the media database.

l l l l

Provides an RPC-based service to manage all the device operations that the nsrmmd process handles on behalf of the nsrd process on the NetWorker server.

Ensures that the necessary device operations are actually performed when needed by nsrd.

Automatically run by nsrd as required.

Only one nsrsnmd runs on each storage node that has configured and enabled devices.

l l l

Provides a uniform library interface to the

NetWorker media management daemon, nsrmmgd.

Manages the library subsystem media, slot, drive, and port resources providing control to move and access the resources within the library subsystems.

One nsrlcpd starts for each configured tape library.

Processes on NetWorker hosts 31

Overview

NMC server

The following table summarizes the processes that support the NMC server software.

Table 5 NMC server processes

Process

nsrexecd gstd httpd postgres gstsnmptrapd

Function

Authenticates and processes the NetWorker server remote execution requests and runs the save and savefs programs on the client.

Known as the Generic Services Toolkit (GST), controls other services that are provided by the NMC server.

Starts the NMC Console GUI on the client through a web browser.

A database server that manages information pertaining to

NMC server management. For example, Console reports.

l l l

Monitors SNMP Traps on a managed Data Domain system.

Provides the ability to report SNMP Trap events in the

NMC Events task.

Started only when SNMP Trap monitoring is configured for the Data Domain system.

Stop and start the NMC server

To complete some tasks in the NetWorker software, stop and start the NetWorker

Console service.

Stopping the NMC server on Windows

Perform the following steps as a Windows administrator to stop the NMC server service, which also stops the postgres and httpd processes.

Procedure

1. Right-click My Computer, and then select Manage.

2. Expand Services and Applications, and then select Services.

3. Right-click EMC GST Service and select Stop.

Note

The EMC GST Service stops the EMC GST Database Service and the EMC GST

Web Service.

32 EMC NetWorker 9.1 Administration Guide

Overview

Starting the NMC server on Windows

Perform the following steps as a Windows administrator to start the NMC server service, which also starts the postgres and httpd processes.

Procedure

1. Right-click My Computer, and then select Manage.

2. Expand Services and Applications, and then select Services.

3. Verify that the NetWorker client is running.

The NetWorker Remote Exec Service should have a status of Started. If the service has not started: a. Right-click NetWorker Remote Exec Service.

b. Select Start.

4. Right-click EMC GST Service, then select Start.

Note

The EMC GST Service starts the EMC GST Database Service and the EMC

GST Web Service.

Stopping the NMC server on Linux

Perform the following steps as root on the NMC server to stop the NMC server process, which also stops the postgres and httpd processes.

Procedure

1. To stop the NMC server processes, type

/etc/init.d/gst stop

2. To confirm that the gstd, httpd, and postgres process are not running, type

ps -ef | grep lgtonmc

Starting the NMC server processes on Linux

Perform the following steps as root on the NMC server to start the NMC process, which also starts the postgres and httpd processes.

Procedure

1. To verify that the NetWorker client process, nsrexecd is running, type

ps -ef

| grep /usr/sbin/nsr

When the client process is running, a message similar to the following appears: root 240 1 0 ? 0:04 /usr/sbin/nsrexecd -s mysrvr

If nsrexecd is not running, type

/etc/init.d/networker start

to start the process.

2. To start the NMC server daemon, postgres, and httpd processes., type

/etc/init.d/gst start

3. To confirm that the gstd, postgres, and httpd processes have started, type

ps -ef | grep lgtonmc

When the processes have started, output similar to the following appears: root 3064 1 0 10:03 ? 00:00:01 /opt/lgtonmc/bin/gstd dbuser 3329 1 0 10:04 ? 00:00:00 /opt/lgtonmc/

Stop and start the NMC server 33

Overview postgres/bin/postgres -D /opt/lgtonmc/nmcdb/pgdata root 3969 1 0 10:04 ? 00:00:00 /opt/lgtonmc/apache/bin/ httpd -f /opt/lgtonmc/apache/conf/httpd.conf

nobody 3970 3969 0 10:04 ? 00:00:00 /opt/lgtonmc/ apache/bin/httpd -f /opt/lgtonmc/apache/conf/httpd.conf

Stop and start a NetWorker server, client, or storage node

This section describes how to manually stop and start the services for a NetWorker server, client, or storage node. In NetWorker 8.0 and later, new attributes have been introduced to configure a NetWorker server to not accept any new backup or recover sessions in preparation of a NetWorker daemon shutdown or server restart.

EMC NetWorker Security Configuration Guide provides more information around how to prevent the NetWorker server from accepting new backup and recover sessions

Stopping a NetWorker host on Windows

Perform the following steps as a Windows administrator to stop the services on a

NetWorker server, storage node, and client.

Procedure

1. Right-click My Computer, and then select Manage.

2. Expand Services and Applications, and then select Services.

3. Right-click NetWorker Remote Exec Service, and then select Stop.

Note

On a NetWorker server, the NetWorker Remote Exec Service stops the

NetWorker Backup and Recovery and the NetWorker Message Queue

Adaptor services. On an NMC server, the NetWorker Remote Exec Service also stops the EMC GST Service.

Starting a NetWorker host on Windows

Perform the following steps as a Windows administrator to start the services on a

NetWorker server, storage node, and client.

Procedure

1. Right-click My Computer, and then select Manage.

2. Expand Services and Applications, and then select Services.

3. Start the appropriate service: l

NetWorker server: Right-click the NetWorker Backup and Recover Server service and select Start.

Note

The NetWorker Backup and Recover Server service also starts the

NetWorker Remote Exec Service and the NetWorker Message Queue

Adaptor service.

NetWorker client or storage node: Right-click the NetWorker Remote Exec

Service and select Start.

34 EMC NetWorker 9.1 Administration Guide

Overview

Stopping a NetWorker host on UNIX

Perform the following steps as the root user to stop the NetWorker processes on a

NetWorker server, storage node, or client.

Procedure

1. To stop the NetWorker processes, type the following command from a prompt:

/etc/init.d/networker stop

2. To confirm that the NetWorker processes are not running, type the following command from a prompt:

ps -ef | grep /usr/sbin/nsr

Starting a NetWorker host on UNIX

Perform the following steps as the root user to start the NetWorker processes on a

NetWorker server, storage node, or client.

Procedure

1. Type the appropriate startup command for the operating system, as summarized in the following table.

Table 6 NetWorker startup commands

Operating system

Solaris, Linux

HP-UX

AIX

Startup command

/etc/init.d/networker start

/sbin/init.d/networker start

/etc/rc.nsr

2. Type

/etc/init.d/networker status

to confirm that the NetWorker processes that are appropriate to the NetWorker installation type have started.

Processes on NetWorker hosts

on page 29 provides more information.

Stopping the NetWorker processes on Mac OS X

Perform the following steps as a Mac Administrator to stop the NetWorker processes on a Mac OS X host.

Procedure

1. Open the Mac OS-X Terminal application utility.

2. To stop the NetWorker processes, type the following command:

launchctl unload /Library/LaunchDaemons/com.emc.NetWorker.plist

Note

The launchd daemon/agent manager controls the NetWorker processes, and

NetWorker configures the processes to run continuously on the host in the background. It is not recommended that you manually stop and start NetWorker processes under normal operating conditions.

Stop and start a NetWorker server, client, or storage node 35

Overview

Starting the NetWorker process on Mac OS X

Perform the following steps as a Mac Administrator to start the NetWorker processes on a Mac OS X host.

Procedure

1. Open the Mac OS X Terminal application utility.

2. Type

launchctl load /Library/LaunchDaemons/ com.emc.NetWorker.plist

to start the NetWorker client process.

NetWorker user interfaces

The NetWorker application consists of several user interfaces that provide the ability to configure and use NetWorker features and functionality.

NMC user interface

The NMC server uses httpd to provide administrators with a graphical user interface to connect to an NMC server and managed NetWorker servers. The NMC UI can be accessed from any computer in the environment with a supported web browser and

Java Runtime Environment (JRE).

The EMC NetWorker Installation Guide provides more information about the web browser and JRE requirements for a host that runs the NMC UI. Multiple users can use different browser sessions on different hosts to access the NMC UI simultaneously.

NMC GUI

Use the NMC GUI to manage an NMC server and NetWorker servers.

The following figure illustrates the NMC GUI.

Figure 2 NMC GUI window

36 EMC NetWorker 9.1 Administration Guide

Overview

Figure 2 NMC GUI window (continued)

The NMC window is the first point of access for NMC and NetWorker tasks. The following table lists the task-based windows that can be opened from the NMC window taskbar.

Table 7 Windows opened from the NMC GUI

Button Window

Enterprise

Reports

Description

Select a NetWorker server to manage and monitor the server and its backup clients.

The

Enterprise

window provides the ability to open the

Administration

window for a NetWorker server.

Configure and view NMC reports.

Setup

Control administrative functions: l l

User management —

Add, edit, and delete

NMC user accounts, restrict user views of servers. The EMC

NetWorker Security

Configuration Guide provides information about user management.

License management —

Manage NetWorker licenses. The EMC

NetWorker Licensing Guide provides information about license management.

NetWorker Administration window

The NetWorker Administration window provides you with the ability to manage and configure NetWorker server resources in a GUI. The NMC UI provides you with the ability to open up a NetWorker Administration window for each managed NetWorker server.

NetWorker client interface

Manual back up, recovery, and archive operations can be performed from a client.

Manual operations are not scheduled. They are client-initiated tasks that are

NetWorker Administration window 37

Overview performed when a user wants to back up, recover, or archive one or more files on the

NetWorker host immediately. You can schedule backup, recovery, and archive operations in the NMC GUI.

On Windows hosts only, you can use the NetWorker User GUI to perform manual back up, recovery, and archive operations.

On UNIX and Windows hosts, you can use command line utilities to perform manual operations: l l l

Use the save command to perform a manual backup.

Use the recover command to perform a manual recovery.

Use the nsarchive command to perform a manual archive.

The EMC NetWorker Command Reference Guide or the UNIX man pages provide more information about these commands.

NetWorker character-based interface

Use the NetWorker character-based interface (nsradmin) to perform configuration and management tasks in the NetWorker server resource database (resdb) and the

NetWorker client resource database (nsrexec).

You can start the nsradmin interface by using this command: nsradmin

For more information about nsradmin, the EMC NetWorker Command Reference

Guide or the UNIX man pages provides more information.

NetWorker command-line interface

Perform client and server tasks by typing commands at the prompt. The EMC

NetWorker Command Reference Guide or the UNIX man pages provides information about these commands.

38 EMC NetWorker 9.1 Administration Guide

CHAPTER 2 Getting Started

This chapter contains the following topics: l l l

NetWorker Management Console interface

....................................................... 40

Connecting to the Administration window

..........................................................44

Getting started with a new installation

...............................................................60

Getting Started 39

Getting Started

NetWorker Management Console interface

The interface for NetWorker Management Console (NMC), also called the NetWorker

Console, consists of both the Console window, and the Administration window.

Connecting to the Console window

The following sections describe how to connect to the Console window:

Before you connect

Ensure that you configure the NetWorker datazone correctly, and that the required daemons are running on the NetWorker server and the NMC server.

Linux NetWorker Server—Confirm that the NetWorker daemons have started, by typing the following command:

/etc/init.d/networker status

. For a

NetWorker server, the nsrctld daemon starts. The nsrctld daemon starts other processes that the NetWorker server requires. Output similar to the following appears when the daemons are started: l

+--o nsrctld (29021)

+--o epmd (29029)

+--o rabbitmq-server (29034)

+--o beam (29038)

+--o inet_gethost (29144)

+--o inet_gethost (29145)

+--o jsvc (29108)

+--o jsvc (29114)

+--o nsrd (29123)

+--o java (29135)

+--o nsrmmdbd (29828)

+--o nsrindexd (29842)

+--o nsrdispd (29853)

+--o nsrjobd (29860)

+--o nsrvmwsd (29968)

+--o connectemc (29131)

+--o eventservice.ru (29154)

+--o jsvc (29158)

+--o jsvc (29159)

+--o java (29838)

+--o node-linux-x64- (29885)

+--o nsrexecd (29004)

+--o nsrlogd (29899)

+--o nsrsnmd (30038)

Linux NMC Server:

1. Type ps -ef | /usr/sbin/nsrexecd. Output similar to the following should appear: root 24959 1 1 13:29 ? 00:00:00 /usr/sbin/nsrexecd

If you do not see this output, type /etc/init.d/networker start.

2. Type

ps -ef | grep lgtonmc

. Output similar to the following should appear: root 3064 1 0 10:03 ? 00:00:01 /opt/lgtonmc/bin/gstd dbuser 3329 1 0 10:04 ? 00:00:00 /opt/lgtonmc/

40 EMC NetWorker 9.1 Administration Guide

Getting Started l l postgres/bin/postgres -D /opt/lgtonmc/nmcdb/pgdata root 3969 1 0 10:04 ? 00:00:00 /opt/lgtonmc/apache/bin/ httpd -f /opt/lgtonmc/apache/conf/httpd.conf

nobody 3970 3969 0 10:04 ? 00:00:00 /opt/lgtonmc/ apache/bin/httpd -f /opt/lgtonmc/apache/conf/httpd.conf

Windows NetWorker Server:

1. Confirm that the following services are started: NetWorker Backup and

Recover Server, NetWorker Message Queue Adaptor, and NetWorker Remote

Exec Service.

2. If these services are not started, start the NetWorker Backup and Recover

Server Service.

Windows NMC Server:

1. Confirm that the following services are started: EMC GST Database Service,

EMC GST Service, and EMC GST Web Service.

2. If these services are not started, start the EMC GST service.

Connecting to the NMC server GUI

Complete the following procedure to connect to the NMC Server GUI from an NMC client. By default, the NetWorker Authentication Service uses the local user database for user authentication. Specify the NetWorker Authentication Service administrator account to log in to the NMC Server. The EMC NetWorker Security Configuration Guide describes how to configure the NetWorker Authentication Service to use LDAP or AD for user authentication.

Procedure

1. From a supported web browser session, type the URL of the NMC Server:

http://server_name:http_service_port

where: l l server_name is the name of the NMC Server.

http_service_port is the port for the embedded HTTP server. The default

HTTP port is 9000.

For example:

http://houston:9000

The gconsole.jnlp file downloads to the host. When the download completes, open the file.

2. When you use Mozilla Firefox on Windows, and the jnlp extension is not associated with Java, you are prompted to choose the program that opens the jnlp file. In the dialog box that appears, select Open with, and then select

Java (TM) Web Start Launcher. If this application does not appear, browse to the Java 8 folder and select the javaws.exe file.

The following figure provides an example of the file association dialog box that appears with the Mozilla Firefox browser.

Connecting to the Console window 41

Getting Started

Figure 3 Associating a jnlp file with Java (TM) web Start Launcher for Mozilla Firefox

3. In the Welcome page, click Start.

Note

If the Start button does not appear but you see a warning message that states that Java Runtime Environment cannot be detected, click the here hyperlink.

4. When you use Internet Explorer, if a security warning appears, select I accept

the risks and want to run this application, then click Run.

5. In the Log in page, specify the NetWorker Authentication Service administrator username and password, and then click OK.

6. In the Licensing Agreement page, select Accept.

7. In the Welcome to the NMC Server Configuration Wizard page, click Next.

8. In the Set authentication server service account for the NMC server page, review the setting and click Next.

9. In the Specify a list of managed NetWorker Servers page: a. Specify the names of the NetWorker Servers that the NMC Server will manage, one name per line.

Note

If the NMC Server is also the NetWorker Server, specify the name of the

NetWorker Server.

b. Leave the default options Capture Events and Gather Reporting Data enabled.

Consider the following option: l

To allow the NMC Server to monitor and record alerts for events that occur on the NetWorker Server, enable the Capture Events option.

EMC NetWorker 9.1 Administration Guide

Getting Started l

To allow the NMC Server to collect data about the NetWorker Server and generate reports, enable the Gather Reporting Data option.

10. Click Finish. The installation starts the default web browser and connects to the NMC server. The NetWorker Management Console and Getting Started windows appear.

11. In the Enterprise window, right-click the NetWorker Server, and then select

Launch Application.

Note

If you do not specify any NetWorker Servers in the Specify a list of managed

NetWorker servers window, the NMC Enterprise window does not display any

NetWorker Servers. To add a host, right-click Enterprise in the left navigation pane and click New

Host. The Add New Host wizard appears.

Connecting to the NMC server after the first time

Use one of the following methods to connect to the NMC server after the initial connection.

l l

Point the browser to the same URL.

Double-click the NMC product name in the Java Web Start Application Manager.

Double-click the desktop button , if one was configured by using the Java Web

Start Application Manager.

Connecting to the NMC GUI using an ssh connection

You can use ssh port forwarding to connect to the NMC server and generate reports, from the NMC client.

Perform the following steps on the NMC client.

Procedure

1. Open an ssh connection from the NMC client to the NMC server with ssh tunnels for ports 9000 and 9001.

For example: ssh -L9000:localhost:9000 -L9001:localhost:9001 -

L5432:localhost:5432 Console_servername -N

Note

If you changed the default NMC server ports, specify the correct port numbers.

2. Use javaws to connect to the NMC server.

For example: javaws http://localhost:9000/gconsole.jnlp

Connecting to the Console window 43

Getting Started

Connecting to the Administration window

The following sections describe how to connect to the Administration window and browse through the interface.

Opening the Administration window

You can add and select a NetWorker server and open the Administration window.

Procedure

1. From the Console window, click Enterprise.

2. Add one or more NetWorker servers: a. Highlight Enterprise in the navigation tree.

b. From the File menu, select New>Host.

c. Type the name of the host on which the NetWorker server is running, and click Next.

d. Select NetWorker for the type of application to be managed.

e. Click Finish.

f. Repeat for all NetWorker servers in the network.

3. From the left pane, click a host in the Enterprise list.

4. From the right pane, click the application and select Enterprise

Launch

Application, or double-click the application. The Administration window opens as a separate application.

Administration window

NetWorker servers are managed through the Administration window.

The following figure illustrates the NetWorker Administration window.

44 EMC NetWorker 9.1 Administration Guide

Figure 4 Administration window

Getting Started

You can toggle between the Administration window and the NMC UI.

The following table lists the windows that can be launched from the Administration window taskbar.

Table 8 Windows that are launched from the Administration window

Button Window

Monitoring

Protection

Recover

Description

Monitor various activities that are related to the NetWorker server. For example, you can monitor the progress of a policy and view any alerts. A portion of the

Monitoring

window always appears at the bottom of the

Administration

window, providing information on Log messages and Alerts.

Manage NetWorker server resources such as clients, groups, policies, probes, and schedules. Provide the ability to monitor, start, stop, and restart data protection policies.

Manage to recover configurations and schedule recover jobs for NetWorker hosts from a centralized location on the NMC server.

Administration window 45

Getting Started

Table 8 Windows that are launched from the Administration window (continued)

Button Window

Devices

Media

Hosts

Server

Description

Add, configure, and operate single or multiple devices, libraries, and silos for the

NetWorker server.

Manage the activities and the resources that are related to backup volumes. For example, you can mount a backup volume or create a label template for backup volumes.

View information about known NetWorker hosts such as the NetWorker version,

CPU type, and operating system. Manage the

NetWorker client resource database. Perform software upgrades on NetWorker hosts by using client push.

Manage NetWorker server resources such as licenses, notifications, user groups, directives, and restricted datazones.

Editing multiple resources

In the NMC Protection window, you can edit an attribute for multiple resources at the same time.

For example, if you want the schedule for all clients within a group to change from the default to “Full Every Friday”, perform the following steps:

Procedure

1. Select each client resource row in the window.

2. Place the cursor in the column you want to change (in this case, the Schedule column).

The color of the column changes when the cursor is in the column.

3. Right-click in that column and select from the list of available options. The options include Edit, Add to, and Remove from, depending on the column selected.

Only the columns that appear in the window can be selected for multiple resource editing. To add a column that is not currently in view: a. Right-click a table header and select Add Column from the drop-down.

b. Select from the list of available attributes.

46 EMC NetWorker 9.1 Administration Guide

Getting Started

Drag-and-drop functionality

Drag-and-drop functionality is available in the Console and Administration interfaces for many tasks.

Drag-and-drop between resource types in the Console window

The drag-and-drop functionality allows multiple resources to be selected and moved from one resource type to another.

In the Enterprise window from the Console interface, you can drag-and-drop to perform the following actions: l l l l l l

Copy an individual folder in the enterprise hierarchy by selecting the folder, press and holding the Ctrl key, and dragging the folder to a new location.

Move an individual folder in the enterprise hierarchy to a new location by selecting and dragging a folder to a new location.

Copy an individual host node in the enterprise hierarchy by selecting and dragging the host to a new parent folder.

Move an individual host node in the enterprise hierarchy by selecting and dragging the host to a new parent folder.

Copy a selected number of objects in a folder to a new folder in the hierarchy tree or folder contents table. Select an individual folder in the navigation tree to display the contents of the folder, select the contents, while pressing Ctrl, drag the contents to a new folder. Select a collection of folders or hosts and drag them to a new folder by creating a copy of the selected contents in a new location.

Move a selected number of objects in a folder to a new folder in the hierarchy tree or folder contents table. Select an individual folder in the navigation tree to display the contents of the folder, select the contents, and drag the contents to a new folder. Select a collection of folders and or hosts and drag them to a new folder by moving the selected contents to a new location.

Note

Only one object may be selected for drag-and-drop in the navigation tree.

Client and group management in the Administration window

The drag-and-drop functionality allows multiple clients or groups to be selected and moved from one location to another. You can use drag-and-drop functionality in the

Protection window to do the following: l l

Copy selected clients to a new NetWorker group:

1. In the left navigation pane, expand the server resource, and then expand the

Groups resource.

2. Select Clients in the directory tree.

3. Drag-and-drop the client objects from the Client Summary table to a group in the directory tree.

Move selected clients from one NetWorker group to another group:

1. Select a group in the directory tree.

Drag-and-drop functionality 47

Getting Started

2. Move clients from the Client Summary table to another NetWorker group.

Library operations in the Devices window

The drag-and-drop functionality allows multiple slots or devices to be managed in the

Devices window.

You can use drag-and-drop functionality to manage media from the Library window from the Devices task, for instance: l l l l

Mount an individual volume onto a device by selecting a slot in the Slots table and dragging it to a device in the Devices table.

Mount multiple volumes to available devices as assigned by the NetWorker server.

To mount multiple volumes, select multiple slots in the Slots table and drag them anywhere in the Devices table.

Unmount a volume from a selected device and deposit it back in its designated slot by selecting an individual device from the Devices table and dragging it anywhere in the Slots table. The volume image displays in the corresponding slot.

Unmount multiple volumes from a selected device and deposit them back in their designated slot by selecting the devices from the Devices table and dragging them anywhere in the Slots table. The volumes display in the corresponding slots.

Copy and paste tabular information to operating system clipboard

Tabular information can be selected and moved to an operating system clipboard by using drag-and-drop functionality. All tables support selection of multiple rows in a table and the ability to copy and paste the data in the selected rows to the system clipboard. Subsequently, the data in the operating system clipboard can be moved to a target application.

Note

Drag-and-drop operations from the operating system clipboard to a table are not supported.

Multiple library devices and slots

A single operation can be performed on multiple library devices and slots. Multiple rows can be selected in both the Devices and Slots tables simultaneously.

In the Devices table for a library, multiple devices can be selected to perform the following operations: l l l

Unmount

Release device (STL only)

Enable/Disable

In the Slots table for a device, multiple volume operations can be performed for the following operations: l l l l l

Mount

Load without mount

Withdraw

Label

Inventory

48 EMC NetWorker 9.1 Administration Guide

Getting Started l

Remove (STL only)

Setting user interaction preferences

Depending on the window button that was selected from the Console window, you can set various user preferences such as the user interface font, font size, parallel windows, and table settings. For the Reports window, there are ways you can enhance the viewing of displayed reports.

Procedure

1. On the main menu, select View.

2. Set the various options available under the selected window button. You may need to click OK, depending on the option selection.

Monitoring NetWorker server activities in the Administration window

The Monitoring window in the NetWorker Administration application enables you to monitor the activities of an individual NetWorker server.

The Monitoring window provides the following types of activity and status information: l l l l

Data protection policies, workflows, and individual actions.

Cloning, recovering, synthetic full backups, and browsing of client file indexes.

Operations that are related to devices and jukeboxes.

Alerts and log messages.

You can also perform some management operations from the Monitoring window, for example, starting, stopping, or restarting a data protection policy.

Procedure

1. From the NMC Console window, click Enterprise.

2. In the Enterprise view, right-click the NetWorker server and select Launch

Application.

The Administration window appears.

3. Click Monitoring to view the Monitoring window.

Setting user interaction preferences 49

Getting Started

Figure 5 Monitoring window

About the Monitoring window

On the Administration window taskbar, select Monitoring to view the details of current NetWorker server activities and status, such as: l l l

Policies and actions.

Cloning, recovering, synthetic backups, checkpoint restart backups, and browsing of client file indexes.

Alerts and log messages, and operations that are related to devices and jukeboxes.

While the Monitoring window is used primarily to monitor NetWorker server activities, it can also be used to perform certain operations. These operations include starting, stopping, or restarting a workflow.

The Monitoring window includes a docking panel that displays specific types of information. Select the types of information you want to view from the docking panel.

A portion of the Monitoring window, which is known as the task monitoring area, is always visible across all windows. A splitter separates the task monitoring area from the rest of the window. You can click and move the splitter to resize the task monitoring area. The arrow icon in the upper right corner of the Monitoring window allows you to select which tasks you want to appear in this view.

50 EMC NetWorker 9.1 Administration Guide

Getting Started floating windows that can be viewed simultaneously. The following table describes the various types of information available in the docking panel, and the details each one provides.

Table 9 Monitoring window panel

Window

Policies

Actions

Sessions

Alerts

Devices

Operations

Log

Information provided

The

Policies

tab provides you with status information about all configure policies and the associated workflows and actions. The

Actions

tab provides you with status information for all actions.

Policies/Actions pane

on page 691 provides more information.

Allows you to customize whether to display all session types, or only certain session types. The information that is provided depends on which session type you select. For example, if you select

Save Sessions

, the window lists clients, save sets, groups, backup level, backup start time, duration of the

backup, devices, rate, and size. Sessions pane provides more

information.

Lists the priority, category, time, and message of any alerts.

Alerts pane

provides more information.

Lists devices, device status, storage nodes, libraries, volumes, pools, and related messages.

Devices pane provides more

information.

Lists the status of all library and silo operations, including nsrjb operations that are run from the command prompt.

Also lists user input, libraries, origin, operation data, operation start time, duration of the operation, progress messages, and error messages.

When displaying Show Details from the

Operations

window, the length of time that the window is displayed depends on the value that is typed in the

Operation Lifespan

attribute on the

Timers

tab of the

Properties

dialog box for the corresponding library. To access library properties, click

Devices

in the taskbar. By default, this pane is hidden.

Lists messages that are generated by the NetWorker server, including the priority of each message, the time the message was generated, the source of the message, and the category.

Log pane

provides more information.

Sessions window

Use the Sessions window to view the sessions that are running on a NetWorker server. You can change the view of this window to display these sessions:

The Sessions pane below the Policies/Actions pane provides details on individual save, recover, clone, and synthetic full sessions by client.

Click the tabs at the bottom of the Sessions pane to view all sessions or to limit the list of sessions by the session type. Session types include:

Monitoring NetWorker server activities in the Administration window 51

Getting Started l l l l

Save

Recover

Clone

Browse l l

Synthetic Full/Rehydrated Sessions

All

To change the displayed session types go to View

Show, and select the type of sessions to display. To display all sessions currently running on the NetWorker Server, regardless of type, select All Sessions.

You can stop a session (backup, synthetic full backup, clone, and recovery sessions) from the Monitoring window, even if the session was started by running the savegrp command.

To stop a session, right-click the session in the pane, and select Stop from the dropdown.

Alerts pane

The Alerts pane displays alerts that are generated by a particular NetWorker server or

Data Domain system that has devices that are configured on the NetWorker server.

The Alerts pane includes priority, category, time, and message information.

An icon represents the priority of the alert. The following table lists and describes each icon.

Table 10 Alerts window icons

Icon Label

Alert

Critical

Emergency

Information

Notification

Description

Error condition detected by the NetWorker server that should be fixed by a qualified operator.

Severe error condition that demands immediate attention.

Condition exists that could cause NetWorker software to fail unless corrected immediately.

This icon represents the highest priority.

Information about the current state of the server. This icon represents the lowest priority.

Important information.

Waiting

Warning

The NetWorker server is waiting for an operator to perform a task, such as mounting a tape.

A non-fatal error has occurred.

When items on the Alerts pane are sorted by the Priority column, they are sorted in alphabetical order based on the label of the icon.

52 EMC NetWorker 9.1 Administration Guide

Devices pane

Getting Started

Removing alerts

Remove individual alert messages from the Events tables by removing them from the

Events table. To delete a message in the Events table, right-click the message, and select Dismiss.

Note

The alert message remains in the Log window in the NetWorker Administration program.

The Devices pane allows you to monitor the status of all devices, including NDMP devices. If the NetWorker server uses shared and logical devices, the window is adjusted dynamically to present a set of columns appropriate for the current configuration.

The Devices pane provides the following information: l l l l l l l l

Status of the operation.

Name of the device.

Name of the storage node that contains the device.

For tape devices, the name of the library that contains the device.

Name of the volume in the device.

Name of the pool that is associated with the volume.

Last message generated for the device.

Whether the operation requires user input.

For example, a labeling operation may want the user to acknowledge whether the system should overwrite the label on a tape.

Entering user input on page 55 provides instructions on how to deal with a user

input notification.

If the current server configuration includes a shared device, a Shared Device Name column appears on the Devices pane. The name of the shared device appears in the

Shared Device Name column. If other devices for that configuration are not shared devices, then the Shared Device Name column is blank for those devices. Only a single device per hardware ID can be active at any particular moment. The information for inactive shared devices is filtered out, and as a result, only one device per hardware ID is presented on the window at any time.

An icon represents the device status. The following table lists and describes each icon.

Table 11 Devices status icons

Icon Label

Library device active

Description

The library device is active.

Library device disabled

Library device idle

The library device is disabled.

The library device is idle.

Monitoring NetWorker server activities in the Administration window 53

Getting Started

Table 11 Devices status icons (continued)

Icon Label Description

Stand-alone device active The stand-alone device is active.

Stand-alone device disabled The stand-alone device is disabled.

Stand-alone device idle The stand-alone device is idle.

When you sort items in the Devices pane by the Status column, NetWorker sorts the devices in alphabetical order based on the label name of the icon.

Operations window

The Operations window displays information about device operations. It provides the following information: l l l l l l l l l

Status of the operation.

Name of the library.

Whether the operation requires user input.

For example, a labeling operation may want the user to acknowledge whether the

system should overwrite the label on a tape. Entering user input

on page 55 provides instructions on how to deal with a user input notification.

The origin, or source, of the operation.

For example, the interface, nsrjb or the NetWorker server.

Time the operation started.

Type of operation.

Duration of the operation.

Status messages from the operation.

Any error messages.

NOTICE

Only the last error message of the operation appears in the Error Messages column. Move the mouse pointer over the cell containing the last error message to display the entire list of error messages.

The operation status is represented by an icon. The following table lists and describes each of the icons.

Table 12 Operations window icons

Icon Label

Failed

Description

The operation failed.

Queued

Retry

The operation is waiting in the queue to run.

The operation failed, but may work if you try again.

54 EMC NetWorker 9.1 Administration Guide

Getting Started

Table 12 Operations window icons (continued)

Icon Label

Running

Description

The operation is running.

Successful The operation completed successfully.

User Input The operation requires user input.

When items on the Operations window are sorted by the Status column, they are sorted in alphabetical order based on the label of the icon.

Viewing operation details

The Operation Details dialog box opens, providing information about the completion of the operation. The Completion Time displays the time that the operation finished.

The time that it took to complete the operation is the difference between the completion and start times of the operation.

To save operation details to a file, click Save in the Operation Details dialog box.

When prompted, identify a name and location for the file.

Procedure

1. From the Administration window, click Monitoring.

2. Click Operations in the docking panel.

3. Right-click the operation, then select Show Details.

Stopping an operation

Certain operations can be stopped from the Operations window.

Procedure

1. From the Administration window, click Monitoring.

2. Click Operations in the docking panel.

3. Right-click the operation to stop, then select Stop.

4. Click Yes to confirm the stop.

Note

Operations that were started from a command line program, such as the nsrjb command, cannot be stopped from the Operations window. To stop these operations, press

Ctrl-c

from the window where the command was started.

Entering user input

If the system requires user input, select the labeling operation in slow/verbose mode and the Supply User Input icon appears.

Procedure

1. Right-click the operation, then select Supply Input.

2. Confirm the requirement to supply input.

Monitoring NetWorker server activities in the Administration window 55

Getting Started l

If Yes, and input is supplied, the icon in the User Input column disappears.

Note

If two users try to respond to the same user input prompt, the input of the first user takes precedence, and the second user receives an error message.

If No, and input is not supplied, the operation will time out and fail.

Log window

To view the most recent notification logs, click the Log window from the docking panel in the Monitoring window. The Log window provides the priority, time, source, category, and message for each log.

Note

If a particular log file is no longer available, check the log file on the NetWorker server.

The log files are located in NetWorker_install_path\logs directory.

An icon represents the priority of the log entry. The following table lists and describes each icon.

Table 13 Icons in the Log pane

Icon Label

Alert

Critical

Emergency

Information

Notification

Description

Error condition that is detected by the

NetWorker server that should be fixed by a qualified operator.

Severe error condition that demands immediate attention.

Condition exists that could cause NetWorker software to fail unless corrected immediately.

This icon represents the highest priority.

Information about the current state of the server. This icon represents the lowest priority.

Important information.

Waiting

Warning

The NetWorker server is waiting for an operator to perform a task, such as mounting a tape.

Non-fatal error has occurred.

When you sort items on the Log pane by using the Priority column, NetWorker sorts the icons in alphabetical order based on the name of the label.

Recover window

The Recover window displays information about recover configurations that are created with the NMC Recovery wizard.

You can use this window to:

EMC NetWorker 9.1 Administration Guide

Getting Started l l

Start the NMC Recovery wizard to create recover configurations or modify saved recover configurations.

Identify the status of a recover configuration that is created with the NMC

Recovery wizard.

Start and stop a recover job.

The Recover window is divided into five sections: l l l

Toolbar—The toolbar is hidden by default. To display the recovery toolbar, select

View

Show toolbar

Summary

Configured Recovers l

Currently Running

A splitter separates the Configured Recovers section from Currently running window. You can click and move the splitter to resize these two windows.

The following table shows an example of the Recover window.

Figure 6 Recover window

Recover toolbar

The Recover toolbar provides you with the ability to quickly perform common recover operations. The following table summarizes the function of each toolbar button.

Table 14 Recovery toolbar options

Button Function

Starts the NMC

Recover

wizard to create recover configurations.

Displays the

Properties

window for the saved recover configuration that you selected in the

Configured Recover

window.

Deletes the saved recover configuration that you selected in the

Configured

Recover

window.

Displays online help for the

Recover

window.

Monitoring NetWorker server activities in the Administration window 57

Getting Started

Table 14 Recovery toolbar options (continued)

Button Function

Displays the

Find

window at the bottom of the

Recover

window. The

Find

window allows you to perform keyword searches for messages that appear in the

Logs

window.

Start the recover operation for a selected saved recover configuration. This option is only available for a recover configuration that has a Never run, or

Failed status.

Stop in-progress recover operation that you selected in the

Currently

Running

window.

Note

The Recover toolbar does not appear by default. To display the Recover toolbar, select View

Show toolbar.

Recover Summary

The Recover Summary section displays a high-level overview of recover jobs.

This section includes the following information: l l

Total Recovers—The total number of successful recover jobs.

Since—The number of successful recover jobs since this date.

Configured Recovers

The Configured Recovers window displays a list of saved recover configurations in a tabular format. You can sort the information by column. The Configured Recovers table displays the following information for each saved recover configuration: l l l l l l l l l l l l

Status—The job status of a saved recover configuration.

Name

Source client

Destination client

Recovery list

Recover type—For example, file system or BBB.

Comment

OS—The operating system of the source host.

Recover requestor—The Windows or UNIX account used to create the recover configuration.

Start Time

End Time

Start date

58 EMC NetWorker 9.1 Administration Guide

Table 15 Save recover configuration job status

Icon Description

The last recover attempt failed.

The last recover attempt completed successfully.

The recover job has never run.

The recover job is scheduled to run in the future.

The recover job has expired.

Getting Started

Currently running

The Currently Running window displays a list of in progress recover jobs in a tabular format. You can sort the information by column. The Currently Running table displays the following information for each job: l l l l l l l l l l l l l l l l

Status

Name

Source client

Destination client

Recovery list

Recover type—For example, file system or BBB

Volume

Comment

Device

Size

Total size

% complete

Rate (KB/s)

Start time

Duration

Currently running

Find

The Find section appears along the bottom of the Recover window, after you select the Find button on the Recover toolbar. Find allows you to search for keywords in the

Monitoring NetWorker server activities in the Administration window 59

Getting Started

Configured Recovers window. The following table summarizes the available find options.

Table 16 Find options

Find option

Find

Prev

Highlight All

Sort Selected

Match case

Description

Highlight the first saved recover configuration that contains the specified keyword.

Highlight the previous saved recover configuration that contains the specified keyword.

Highlights each saved recover configuration that contains the specified keyword.

Sorts each highlighted recover configuration in the

Configured Recover table so that they appear at the top of the Configured Recover table.

Make the keyword search case sensitive.

Getting started with a new installation

The following section provides basic information on how to get started with a new installation by configuring the NetWorker datazone and starting the NetWorker

Management Console (NMC) Enterprise window and Administration window.

Common NetWorker tasks

There are several common tasks available in the NetWorker Console.

Adding a new host

You can add hosts by using the NetWorker Console.

Procedure

1. Log in to Console as a NetWorker Administrator.

Click the Enterprise button on the taskbar.

3. Right-click Enterprise in the navigation tree.

4. Select New

Host.

5. In the Host Name field, specify the IP address or DNS name of the NetWorker server and click Next.

6. On the Select Host Type window, select NetWorker and click Next.

7. On the Manage NetWorker window, leave the default options Capture Events and Gather Reporting Data enabled.

l l

Enable the Capture Events option to allow the NMC server to monitor and record alerts for events that occur on the NetWorker server.

Enable the Gather Reporting Data option to allow the NMC server to automatically collect data about the NetWorker server and generate reports on the NMC server.

60 EMC NetWorker 9.1 Administration Guide

Getting Started

8. Click Finish.

Device configuration

You can configure devices to test the NetWorker software.

Configuring a stand-alone tape device

Procedure

1. Log in to the NMC GUI as an administrator of the NetWorker server.

Click the Enterprise button on the taskbar.

3. Highlight a host in the navigation tree: a. Right-click NetWorker.

b. Select Launch Application. The NetWorker Administration window appears.

Click the Devices button on the taskbar.

5. In the navigation tree view, right-click a host and select Scan for Devices.

The Scan for Devices window appears.

6. On the Select Target Storage Nodes window, perform either of the following steps: l

Select the storage node for the library.

Click Create a new Storage Node to create a storage node.

7. Select Start scan.

NetWorker scans for new devices and the Log pane provides the status of the scan operation.

8. On the left pane, select Devices and then from the right pane, select the new device.

9. From the Devices menu, select Devices

Device Operations

Label.

10. In the Label window, verify the information and click OK.

Configuring a stand-alone advanced file type device

Create a device that is local to the NetWorker server to receive the backup data.

Procedure

1. Log in to the NMC GUI as an administrator of the NetWorker server.

Click the Enterprise button on the taskbar.

3. Highlight a host in the navigation tree: a. Right-click NetWorker.

b. Select Launch Application. The NetWorker Administration window appears.

Click the Devices button on the taskbar.

5. From the File menu, select New Device Wizard.

6. On the Select the Device Type window, select Advanced File Type Device

(AFTD), then click Next.

Common NetWorker tasks 61

Getting Started

7. On the Select Storage Node window, leave the default values, and click Next.

8. On the Select the Device Path window, select an empty folder or create a new folder on the NetWorker server, then click Next.

9. On the Configure Device Attributes window, specify a name for the new device in the NetWorker Device Name field, for example:

myaftd

, and click

Next.

10. On the Label and Mount Devices window, leave the default values and click

Next.

11. In the Review the Device Configuration Settings window, review the configuration information, and click Configure.

12. Click Finish.

Configuring an autochanger or silo

You can configure a new library resource.

Procedure

1. Log in to the NMC GUI as an administrator of the NetWorker server.

Click the Enterprise button on the taskbar.

3. Highlight a host in the navigation tree: a. Right-click NetWorker.

b. Select Launch Application. The NetWorker Administration window appears.

Click the Devices button on the taskbar.

5. From the left pane, select Storage Nodes.

6. Right-click the storage node for the device and select Configure All Libraries.

7. On the Provide General Configuration Information window, leave SCSI/

NDMP selected and click Next.

8. On the Select Target Storage Nodes window, perform either of the following steps: l l

Select the storage node for the library.

Click Create a new Storage Node to create a storage node.

9. Click Start Configuration.

NetWorker scans for new devices and the Log pane provides the status of the scan operation.

10. Click Finish.

Labeling media

You can label tapes from the NMC GUI.

Procedure

1. Log in to the NMC GUI as an administrator of the NetWorker server.

Click the Enterprise button on the taskbar.

3. Highlight a host in the navigation tree:

EMC NetWorker 9.1 Administration Guide

Getting Started a. Right-click NetWorker.

b. Select Launch Application. The NetWorker Administration window appears.

Click the Devices button on the taskbar.

5. In the navigation tree view, expand Libraries and highlight a library, or select

Devices.

6. In the Device list, right-click a device and select Label.

Scheduling backups

Perform scheduled backups to automatically backup client data on an ongoing basis.

Data protection policies enable you to define the client resources, schedule, and other settings for the backup. The client resources and backup storage resources must also be configured.

Procedure

1. Configure the backup storage resources: a. Configure the storage node that will own the backup storage devices.

b. Configure the backup storage device.

c. Create a label template for labeling volumes, or use one of the preconfigured label templates.

d. Create media pools for sorting and storing backup data.

Backup Storage on page 69 provides more information on configuring backup

storage resources.

2. Configure one or more client resources for each client computer by using either the Client Backup Configuration Wizard or the Client Properties dialog box.

When you configure a client resource, you specify backup settings for the client, including: l l l l

The save sets for the client, which define the data to back up on the client.

Whether to automatically restart failed backups from a known good point, which is called checkpoint restart.

Whether to bypass the storage node and send backup data directly to AFTD or DD Boost storage devices, which is called Client Direct.

Directives that control how the NetWorker server processes files and directories during the backup.

l l

Probe resources for probe-based backups, where the NetWorker server probes the client for a user-defined script before the backup starts.

Whether to back up each save set for the client by using multiple parallel save streams.

Backup command customizations.

Client resources on page 420 provides more information on configuring client

resources.

3. Configure a data protection policy for scheduled backups: a. Create a group to define the client resources to back up.

Common NetWorker tasks 63

Getting Started

The type of group that you create depends on the type of backup that you are performing: l l l

Create a client group or dynamic client group for a traditional backup or a server backup.

Create a VMware group to back up virtual machines or VMDKs.

Create a NAS device group to perform snapshot backups on NAS devices.

b. Create a policy.

Policies provide a container for the workflows, actions, and groups that support and define the backup.

c. Within the policy, create a workflow.

Workflows define the start time for a series of actions, the order of actions in a sequence, and the group of client resources to back up.

d. Create a backup action.

When you create a backup action, you define the following settings: l l l

The type of backup to perform each day.

The destination storage node and media pool.

The retention setting for the backup, which specifies how long to retain the backup data.

e. (Optional) Create other actions for the workflow.

Actions that you may want to include in a backup workflow include: l l l

Check connectivity to verify connectivity between the NetWorker server and the client computer.

Probe to probe a NetWorker client for a user-defined script before the backup starts.

Clone to automatically clone the save sets that result from the backup.

Data Protection Policies on page 207 provides more information on configuring

groups, policies, workflows, and actions.

Viewing failed backups

You can view the details for failed NetWorker backups.

Procedure

1. Log in to the NMC GUI as an administrator of the NetWorker server.

Click the Enterprise button on the taskbar.

3. Highlight a host in the navigation tree: a. Right-click NetWorker.

b. Select Launch Application. The NetWorker Administration window appears.

Click Monitoring .

The Monitoring window displays four windows panes. The Log pane provides a summary of NetWorker server events. The Policies pane displays all configured

EMC NetWorker 9.1 Administration Guide

Getting Started policies on the NetWorker server. To view details information about the status of the actions in a workflow, expand the policy, right-click the workflow, and select Show Details.

Using nsrlogin for authentication and authorization

When you configure the NetWorker Authentication Service to use LDAP/AD authentication, you modify the External Roles attribute in the User Group resource to assign privileges to LDAP and AD users. As a result, NetWorker command line operations and NetWorker module operations might fail due to insufficient privileges.

To resolve this issue, use the nsrlogin command to contact the NetWorker

Authentication Service and authenticate a user. When user authentication succeeds, the NetWorker Authentication Service issues a token to the NetWorker host for the user, which provides CLI operations with token-based authentication until the token expires.

Before you begin

Ensure that the user that the NetWorker Authentication Service validates has the appropriate User Group privileges to run the CLI commands.

Perform the following steps on a NetWorker client on which you initiate the CLI commands, or the requesting host.

Procedure

1. Use the nsrlogin command to validate a user and generate a token for the user:

nsrlogin [-s NetWorker_server] [-H authentication_host] [-P

port] [-t tenant] [-d logindomain] -u username [-p "password"]

where: l l l l l l l

-s NetWorker_server

—Specifies the name of the NetWorker server. Use this option when you use the nsrlogin command on a NetWorker host that is not the NetWorker server.

-H authentication_host

—Specifies the name of the NetWorker

Authentication Service host. Use this option when you use the nsrlogin command on a NetWorker host that is not the NetWorker server. This option is only required when you do not use the

-s

option.

-P port

—Specifies the NetWorker Authentication Service port number.

Use this option when you do not use the

-s

option and when the NetWorker

Authentication Service does not use the default port number 9090 for communications.

-t tenant

— Specifies the tenant name that the NetWorker Authentication

Service should use to verify the username and password. When you omit this option, NetWorker Authentication Service uses the Default tenant to verify the user credentials.

-d logindomain

—Specifies the domain name that the NetWorker

Authentication Service should use to verify the username and password with an external authentication authority. When you omit this option, the

NetWorker Authentication Service uses the local user database to verify the user credentials.

-u username

—Specifies the username that the NetWorker Authentication

Service should validate to generate a token.

-p "password"

—Specifies the password that the NetWorker

Authentication Service should use to verify the username. If you do not

Common NetWorker tasks 65

Getting Started specify the password, the nsrlogin command prompts you to provide the password.

For example, to generate a token for user Konstantin in the idddomain domain and the idd tenant, type the following command:

nsrlogin -s bu-idd-nwserver2 -d idddomain -u Konstantin -p

"1.Password"

Authentication succeeded.

When NetWorker Authentication Service successfully validates the user, the service issues an authentication token to the requesting host.

2. Type the NetWorker command, at the command prompt.

If the validated user does not have the appropriate privileges to run the command, an error message appears or the command does not return the expected result. For example, when you try to perform an operation with a user account that does not have the required privilege, a message similar to the following appears:

Permission denied, user must have the 'Operate NetWorker' privilege'.

Results

The CLI command uses the authenticated token, until the token expires. By default the token expiration period is 4800 minutes, or 8 hours. When the token expires and the user tries to run a CLI command, the command fails with a permissions error and a message similar to the following appears to indicate that the token has expired:

Security token has expired

To resolve this issue, run the nsrlogin command again to generate a new authenticated token.

Note

To revoke the user token and enable the CLI commands to use the Users attribute in the Usergroups resources to authenticate users, use the nsrlogout command. The nsrlogout UNIX man page and the

EMC NetWorker Command Reference Guide provides detailed information about the nsrlogout command.

Performing a manual backup

Perform a manual backup of a file or folder, to test the NetWorker installation. The procedure to perform a manual backup is different on Windows and UNIX.

Performing a manual backup on Windows

Use the NetWorker User program to perform a manual backup Windows. The

NetWorker User program provides a graphical interface to perform manual backups.

Procedure

1. On a NetWorker client, start the NetWorker User program.

2. In the Change server window, select or type the name of the NetWorker server.

3. In the Source and Destination client windows, select the current NetWorker client.

EMC NetWorker 9.1 Administration Guide

Getting Started

4. Click Backup.

5. In the left pane of the Backup window, click the appropriate directory folder.

6. Select a file or directory file to back up in one of the following methods: l l

Select the directory or file and click Mark. To clear an item, click Unmark.

Right-click the directory or file.

When you mark a directory or file for backup, a check mark appears next to that item.

7. Click Start.

The Backup Status window displays the progress of the backup. When the

NetWorker server has successfully finished the backup, this message appears:

Backup completion time: 2-15-07 3:27p

If the backup fails, then: l l l

Review the NetWorker daemon.raw log file on both the NetWorker server and client hosts. Use the nsr_render_log program to review the log file in a readable format. The EMC NetWorker Command Reference Guide describes how to use the nsr_render_log program.

The location of the daemon.raw file is different on Windows and UNIX: n n

On Windows, the log file appears in the C:\Program Files\EMC

NetWorker\nsr\logs directory.

On UNIX, the log file appears in the /nsr/logs directory.

To determine the cause, refer to the Troubleshooting chapter.

Review the operating system log files (Application event log on a Windows client) for more information.

Performing a manual backup on UNIX

Use the save program to perform a manual backup from the command prompt.

For example, to back up /tmp/myfile.txt to a server called jupiter, type:

save -s jupiter /tmp/myfile.txt

The UNIX man pages describe how to use the save program.

Common NetWorker tasks 67

Getting Started

68 EMC NetWorker 9.1 Administration Guide

CHAPTER 3 Backup Storage

This chapter contains the following topics: l l l l l l l l l

Label templates

..................................................................................................70

Media pools

........................................................................................................ 77

Storage nodes

....................................................................................................93

Disk storage devices

......................................................................................... 102

Libraries and silos

............................................................................................. 130

File type devices

............................................................................................... 190

Stand-alone devices

.......................................................................................... 191

Labeling volumes

...............................................................................................197

Troubleshooting devices and autochangers

...................................................... 198

Backup Storage 69

Backup Storage

Label templates

The NetWorker server creates a unique label for each volume by applying a label template. This section describes how label templates and media pools are used to sort, store, and track data on media volumes.

Using label templates

The NetWorker server selects the media pool to which a given set of data is written. A volume is associated with a media pool by its volume label.

The contents of the volume label follow rules that are defined in a specific label template. You then associate a label template with a specific media pool in the Media

Pool resource. If you do not associate data with a specific media pool, the NetWorker server uses the preconfigured Default media pool and corresponding Default label template.

The following figure illustrates how a media pool configuration uses its associated label template to label a volume. For the label template name to appear as a choice in the

Media Pool resource, you must configure a label template before configuring the associated media pool.

Figure 7 Labeling a volume by using a label template

How the NetWorker server uses volume labels

A volume label is a unique internal code, applied by the NetWorker server, that initializes the volume for the server to use and identifies a storage volume as part of a specific pool.

Using media pools

on page 77 provides more information about pools.

Labeling a volume provides a unique name for tracking and recognizing the media, as well as references to volume labels in the records stored in the media database. The

NetWorker server uses the media database records to determine which volumes are needed for backing up or recovering data.

When NetWorker labels a volume, the label operation performs the following actions:

1. Verifies that the volume is unlabeled.

70 EMC NetWorker 9.1 Administration Guide

Backup Storage

2. Labels the volume with the name specified in the Volume Name attribute by using one of the following: l l

The next sequential label from the label template that is associated with the chosen pool.

If a recyclable volume from the same pool is relabeled, the volume label name and sequence number remain the same, but access to the original data on the volume is destroyed. The volume becomes available for new data.

An override volume name that was entered by the user.

Preconfigured label templates

The NetWorker server contains these preconfigured label templates, which correspond to the preconfigured media pools: l l l l l l l l

Guidelines for completing Label Template attributes

There are certain guidelines to keep in mind when completing the attributes for a Label

Template resource. The following table describes how to complete the key attributes for this resource.

Table 17 Key label template attributes

Attribute

Name

Guidelines

Keep the label name consistent with the media pool name, so that the label name

Preconfigured label templates 71

Backup Storage

Table 17 Key label template attributes (continued)

Attribute

Fields

Separator

Guidelines

reflects how the data is organized. For example, a label template named "AcctFull" would identify volumes that belong to a media pool called "Accounting Full."

Do not use these characters in label template names:

/ \ * ? [ ] ( ) $ ! ^ ; ’ " ’ ~ < > & | { } : - . _

A label template is made up of one or more fields. Each field, or component, provides a layer of specificity to your organizational structure. There can be any number of components, but it is best to keep the template simple with as few as necessary. The label cannot exceed 64 characters.

You can use four types of components: l l l l

Range of numbers (for example, 001-999)

Range of lowercase letters (for example, aa-zz)

Range of uppercase letters (for example,

AA-ZZ)

Character string (for example,

Accounting)

Each range includes a start value, a dash

(-), and an end value. The start value and the end value must have the same number of characters. For example, use 01-99

(not 1-99) or aaa-zzz (not aa-zzz).

The order in which you enter each component of the Field attribute is important.

The NetWorker server applies each component in a left-to-right order, starting with the first one entered.

Choose the symbol to appear between component entries. Use the period, dash, colon, or underscore to separate each component of the label template. If label components do not have separators (for example, AA00aa), the labels can be difficult to read.

Choose the next sequence number to write on the label that the NetWorker server places on a volume (according to the template).

72 EMC NetWorker 9.1 Administration Guide

Backup Storage

Table 17 Key label template attributes (continued)

Attribute Guidelines

l l

To force a label to start the label scheme at a particular point, type a start label value. The server continues to generate labels from that point on, according to the rules of the template.

To have the NetWorker server generate the first label, leave this attribute blank.

When the NetWorker server recycles a storage volume, the volume label does not change as long as the volume remains in the same media pool. That is, if a storage volume labeled "Dev.006" is recycled, it retains the volume label "Dev.006" and does not receive a new label with the next sequence number.

The following table lists examples of number sequences for volume labels.

Table 18 Examples of number sequences for volume labels

Type of components

Range of numbers

Character string

Range of numbers

Range of lowercase letters

Range of numbers

Fields

001-100

SalesFull

001-100 aa-zz

00-99

Number sequence result

001, 002, 003,...100

SalesFull.

001,...SalesFull.100

Total number of labels

100

100 aa.00,...aa.99, ab.00,...ab.99,

: ac.00,...ac.99, az.00...az.99,

: ba.00,...ba.99

zz.00,...zz.99

67,600 (262 times

102)

The label template should allow for expansion of the backup media storage system.

For example, it is better to create a template for 100 tapes and not use all of them, than it is to create a template for only 10 tapes and run out of labels. When the server reaches the end of the template numbering sequence, it wraps to the starting value.

For example, after zz.99 (used for the 67,600th label), the next label the server uses is aa.00 for label 67,601.

Guidelines for completing Label Template attributes 73

Backup Storage

Note

When the NetWorker server recycles a volume, the volume label does not change if the volume remains in the same media pool. That is, if a volume labeled Dev.006 is recycled, it will retain the volume label Dev.006 and will not receive a new label with the next sequence number. The original data on the volume, however, will be overwritten by the new data.

Naming label templates

The NetWorker server is packaged with preconfigured label templates that correspond to the preconfigured media pools. If you choose to create the templates, you can include any number of components in the Fields attribute. However, it is best to keep the template simple with as few components as necessary for your organization.

For example, if you create a label template for an accounting department, you can customize the label template in several ways, depending on the size of the storage system and media device capabilities.

The following table illustrates several ways you can use components to organize labels.

Table 19 Using label template components

Separator Resulting volume labels

Type of organizational structure

Sequential

Components

AcctFull

‘001-100

Storage oriented (for example, 3 storage racks with 5 shelves each, each shelf holding 100 tapes)

1-3

1-5

001-100

Two-sided media (for example, optical devices)

AcctFull

000-999 a-b period dash underscore

AcctFull.001

(100 total labels)

1-1-001

This label is for the first tape in rack 1 on shelf 1.

(1,500 total labels)

AcctFull_000_a (side

AcctFull_000_b (side

(2,000 total labels)

Tips for labelling

Naming schemes vary from site to site. One way is to name the volumes with the name of the NetWorker server followed by a three-digit number, for example: jupiter.001

Consider that the simpler a convention is, the easier it can be understood by operators and administrators.

EMC NetWorker 9.1 Administration Guide

Backup Storage

The maximum length for a volume name is 63 characters. With advanced file type devices (adv_file), the maximum length is 60 characters.

Each volume should have a physical (adhesive) label attached to it. Since the

NetWorker server keeps track of the backups and which volumes they are on, you can name the volumes with any convenient name. For example, you can label your volumes

1, 2, 3, or Monday.1, Tuesday.1, Wednesday.1. You can assign a volume any name as long as each one is unique.

The adhesive label on the volume should match the name generated by NetWorker.

For example, if you physically label a volume mars.1, its NetWorker name should also be mars.1.

Working with label templates

This section explains how to create, edit, copy, and delete label templates.

Creating a label template

When creating a label template, consider the labeling guidelines for the Name, Fields,

Separator, and Next components.

Procedure

1. In the Administration window, click Media.

2. In the expanded left pane, select Label Templates.

3. From the File menu, select New.

4. Enter the components for the label template: l l l l

Name: The name of the new label template.

Comment: Any user-defined description or explanatory remarks about the label.

Fields: A list of label components.

Separator: The character to be inserted between label components. If no symbol is selected, the components will have no separators, such as hostarchive[001-999].

Next: (Optional) Enter the next label to be generated by the template.

5. Click OK.

Editing a label template

You cannot change the name of a label template. However, to change an individual label name, delete the existing name in the Next text box, and type a new name.

Procedure

1. In the Administration window, click Media.

2. In the expanded left pane, select Label Templates.

3. In the right pane, perform one of the following tasks: l l

To modify multiple attributes in a single configuration resource by using the

Label Template Properties window, right-click the staging configuration and select Properties.

To modify a specific attribute that appears in the resource window, place the mouse in the cell that contains the attribute that you want to change, then right-click. The menu displays an option to edit the attribute. For

Working with label templates 75

Backup Storage example, to modify the Comment attribute, right-click the resource in the

Comment cell and select Edit Comment.

Note

To modify a specific attribute for multiple resources, press and hold the Ctrl key, select each resource, and then right-click in the cell that contains the attribute that you want to change. The menu displays an option to edit the attribute.

4. Make any required changes, then click OK.

Copying a label template

Procedure

1. In the Administration window, click Media.

2. In the expanded left pane, select Label Templates.

3. In the right pane, select the label template to copy.

4. From the Edit menu, select Copy. The Create Label Template dialog box appears, containing the same information as the label template that was copied, except Name attribute.

5. In the Name attribute, type the name for the new label template.

6. Edit any other attributes as appropriate, and click OK.

Deleting a label template

You cannot delete a preconfigured label template or a label template that is in use.

Procedure

1. In the Administration window, click Media.

2. In the expanded left pane, select Label Templates.

3. In the right pane, select the label template to delete.

4. From the File menu, select Delete.

5. When prompted, click Yes to confirm the deletion.

Setting up a label template to identify volumes

If you are not using tapes with barcode labels, and the Match Bar Code Labels attribute is not enabled for the Library resource, then every backup volume requires a unique label for identification. The NetWorker server creates a unique label for each volume by applying a label template.

Procedure

1. From the Administration window, click Media.

2. In the expanded left pane, select Label Templates.

3. From the File menu, select New.

4. In the Name attribute, type a name for the label template.

5. In the Comment attribute, type a description for the label template.

6. In the Fields attribute, type the label’s components. Place each label component on a separate line. The template can use any or all of these components, although at least one range component must be added:

76 EMC NetWorker 9.1 Administration Guide

Backup Storage l l l l l

Range of numbers—For example, 001-999

Range of lowercase letters—For example, aa-zz

Range of uppercase letters—For example, AA-ZZ

Character string—For example, Accounting

Ranges of numbers or letters change incrementally with each new label. For example: n n

First label: Accounting.001

Second label: Accounting.002

Third label: Accounting.003

7. Select a Separator and click OK. If no symbol is selected, the components will have no separators (for example, Accounting001).

8. Click OK.

Media pools

NetWorker uses media pools and volume labels to sort backup and clone data on media.

Media is a specific collection of volumes to which the NetWorker server writes data.

For example, a tape volume or a Data Domain device. A volume is identified with a unique label based on user configurable label templates.

Media pools act as filters that tell the NetWorker server which backup volumes should receive specific data. The NetWorker server uses media pools along with label templates to track what data is on which specific volume. When you use a barcodeenabled tape library, the NetWorker server uses media pools along with the volume barcode Labels to track which data is on a specific volume.

Note

NetWorker does not use media pools for backup and clone operations to deduplication devices.

Using media pools

Action resources contain an attribute that defines the media pool to which NetWorker should send the backup or clone data.

When a backup or clone action starts, the NetWorker server then checks if a correctly labeled volume for that media pool is mounted on a storage device. If a correctly labeled volume is mounted on a storage device, the NetWorker server writes data to the volume. If there is no correctly labeled volume mounted on a storage device, the

NetWorker server generates a request to mount a volume that is labeled for the pool, and waits until an operator or an autochanger mounts an appropriate volume.

Media pools 77

Backup Storage

Preconfigured media pools

NetWorker provides you with the following preconfigured media pools.

Table 20 Preconfigured media pools

Pool name

Changes to the Client and Pool resources after migration

NetWorker uses a number of attributes that are defined in multiple resources to determine which pool receives the data that is generated by an action task, and how

NetWorker backs up the data. The migration process preserves the values that are defined for the attributes and introduces new attributes in the Action resource.

NetWorker provides the following attributes, which work together to determine how

NetWorker manages a backup and determines which device to use to receive the backup data: l

Client resource—Pools, Retention, Save set, and Level attributes on the General tab of the Client Properties window. The migration process retains the values in these legacy attributes.

Using media pools 79

Backup Storage

Note

The Modify Client wizard does not display the Pools, Retention, Save set, and

Level attributes.

l l

Action resource—Destination Pool and Retention attributes on the Specify the

Backup Options and Specify the Clone Options wizard windows. The backup levels are defined for the action schedule on the Specify the Action Information wizard window.

Pool resource—Clients, Save sets, and Retention policy attributes on the

Legacy tab. The values that appear in these attributes were defined in NetWorker

8.1.x and 8.2.x. After the migration completes, the NetWorker 9.1 server retains the values and these legacy attributes become read-only. You cannot modify the values in these fields after migration.

The Action resource includes an attribute that is called Client Override Behavior. The value that is selected for this attribute determines which resource attribute has precedence over the attributes in other resources that determine the same behavior.

By default, the migration process enables Legacy Backup Rules on an Action resource. Legacy Backup Rules allow NetWorker to use the values during the pool selection criteria process.

Note

By default, the NetWorker Administration window does not show the legacy attributes. To view the legacy attributes in the Client Properties window, go to the

View menu and select Diagnostic Mode.

Pool selection criteria

EMC recommends that you use the configuration settings in an Action resource to determine which pool received backup data. NetWorker provides you with the ability to configure a Pool attribute in the client resource, which can override the value defined in the Action resource. Additionally, the Pool resource contains 8.2.x legacy attributes that provide you with the ability to define backup data criteria for the pool.

How and when NetWorker uses the attributes values defined in the Pool, Action, and

Client resources to determine which backup pool will receive data depends on the value that you select in the Client Override Behavior attribute of the Action resource: l l l

Client Can Override—The value in Pool attribute of the client resource takes precedence over the Destination pool value that is defined in the Action resource.

NetWorker does not use the values that are defined in the Client, Save set, and

Levels attributes of the Pool resources when deciding which pool receives backup data for a client.

Client Can Not Override—The value defined Destination Pool attribute in the

Action resource takes precedence over the value that is defined in the Pool attribute of the Client resource. NetWorker does not use the values that are defined in the Client, Save set, and Levels attributes of the Pool resources when deciding which pool receives backup data for a client.

Legacy Backup Rules—Enabled for migrations only. NetWorker uses the values that are defined in the Client, Save set, and Levels attributes of the pool resource to determine which pool receives backup data from a client. The values that are defined in the Client, Save set, and Levels of the pool resource take precedence over the Destination Pool value that is defined in the Action resource, and the

Pool value that is defined in the Client resource.

EMC NetWorker 9.1 Administration Guide

Backup Storage

Note

You cannot modify the legacy attributes in the migrated Pool resources.

The following table summarizes how NetWorker determines which pool receives the backup data, which is based on the configuration of the Action, Client, and Pool resource attributes.

Table 21 Determining which pool receives backup data

Client Override

Behavior

(Action)

Destination pool

(Action)

Client Can Override Defined

Pool

(Client)

Defined

Client Can Override Defined

Client Cannot

Override

Legacy Backup

Rules

Legacy Backup

Rules

Legacy Backup

Rules

Defined

Undefined

Defined

Undefined

Defined

Undefined

Legacy criteria attributes (Pool)

Pool that receives the data

Defined and criteria matches

Pool defined in

Client resource

Defined and criteria matches

Pool defined in

Action resource

Defined and criteria matches

Pool defined in

Action resource

Defined and criteria matches

Pool that matches legacy criteria

Defined and criteria matches

Undefined or no matches

Pool that matches legacy criteria

Default

Example 1 Client Can Override is enabled

A Protection group contains two clients,

SQL_clnt and Exchange_clnt. The workflow that is associated with the protection group contains a backup action.

l l l l

Backup action configuration: n

Destination Pool=

App_backups n n

Schedule=Daily full backup

Client Override Behavior=Client Can Override

The Pool attribute that is defined for the

SQL_clnt client resource is

SQL_backups.

The Pool attribute for

Exchange_clnt is Exchange_backups.

The Full level is enabled in the Levels attribute of a pool resource named

Backups.

In this example, NetWorker sends the backup data for

Exchange_clnt to

Exchange_backups, the pool that is defined in the backup action. NetWorker sends the backup data for

SQL_clnt to the pool defined in the client resource,

SQL_backups.

Example 2 Example: Client Can Override is enabled

Using media pools 81

Backup Storage

Example 2 Example: Client Can Override is enabled (continued)

A Protection group contains two clients, SQL_clnt and Exchange_clnt. The workflow that is associated with the protection group contains a backup action.

l l l l

Backup action configuration: n n

Destination Pool= App_backups

Schedule=Daily full backup n

Client Override Behavior=Client Can Override

The Pool attribute that is defined for the

SQL_clnt client resource is

SQL_backups.

The Pool attribute for Exchange_clnt is not defined.

The Full level is enabled in the Levels attribute of a pool resource named

Backups.

In this example, NetWorker sends the backup data for

Exchange_clnt to

App_backups, the pool that is defined in the backup action. NetWorker sends the backup data for

SQL_clnt to the pool defined in the client resource, SQL_backups.

Example 3 Client Cannot Override is enabled

A Protection group contains two clients, SQL_clnt and Exchange_clnt. The workflow that is associated with the protection group contains a backup action.

l l l l

Backup action configuration: n n

Destination Pool= App_backups

Schedule=daily full backup n

Client Override Behavior=Client Cannot Override

The Pool attribute that is defined for the

SQL_clnt client resource is

SQL_backups.

The Pool attribute for Exchange_clnt is Exchange_backups.

The Full level is enabled in the Levels attribute of a pool resource named

Backups.

In this example, NetWorker sends the backup data for

SQL_clnt and Exchange_clnt to

App_backups, the pool that is defined in the backup action.

Example 4 Legacy Backup Rules is enabled

A Protection group contains two clients, SQL_clnt and Exchange_clnt. The workflow that is associated with the protection group contains a backup action.

l l

Backup action configuration: n n

Destination Pool= App_backups

Schedule=daily full backup n

Client Override Behavior= Legacy Backup Rules

The Pool attribute that is defined for the

SQL_clnt client resource is

SQL_backups.

82 EMC NetWorker 9.1 Administration Guide

Backup Storage

Example 4 Legacy Backup Rules is enabled (continued) l l

The Pool attribute for

Exchange_clnt is not defined.

The Full level is enabled in the Levels attribute of a pool resource named Backups.

In this example, NetWorker sends the backup data for SQL_clnt and Exchange_clnt to

Backups, the pool that matches the level Full backup criteria.

Example 5 Legacy Backup Rules is enabled

A Protection group contains two clients,

SQL_clnt and Exchange_clnt. The workflow that is associated with the protection group contains a backup action.

l l l l

Backup action configuration: n n

Destination Pool=

App_backups

Schedule=daily full backup n

Client Override Behavior= Legacy Backup Rules

The Pool attribute that is defined for the SQL_clnt client resource is

SQL_backups.

The Pool attribute for

Exchange_clnt is not defined.

The manual level is enabled in the Levels attribute of a pool resource named

Backups.

In this example, NetWorker sends the backup data for SQL_clnt and Exchange_clnt to the Default pool because a pool does not exist with legacy attributes that match the configuration for the backup data.

Matching the pool criteria with Legacy Backup Rules enabled

After a migration and configuring media pools, data generated by an action might match the criteria for more than one media pool configuration. For example, if you configure one media pool to accept data from a client that is called mnd.emc.com, and you configure another media pool to accept data from all full backups, NetWorker uses other criteria to determine which pool of volumes receives the data from a full backup of the mnd.emc.com client.

The NetWorker server uses the following media pool selection criteria:

1. Groups attribute (highest precedence)

2. Clients attribute

3. Save sets attribute

4. Levels attribute (lowest precedence)

When data matches the attributes for two media pools, for example, Client and Level, the data is written to the media pool specified in the Client attribute. For example, in the case where the data from the client matched the criteria for two different media pools, the data is routed to the media pool that accepts data from the mnd.emc.com

client.

The following table details the hierarchy that the NetWorker server uses to determine media pool selection when a conflict arises. For example, the media pool criteria for

Groups takes precedence over the media pool criteria for Clients, Save sets, and

Levels. If data does not meet the criteria for any customized pool, NetWorker writes the data to the Default media pool.

Using media pools 83

Backup Storage

Table 22 NetWorker hierarchy for resolving media pool conflicts

Precedence

Highest

Groups attribute

x x x x x x x x

Clients attribute

x x x x x x x x

Save sets attribute

x x x x x x x x

Lowest

Working with media pools

This section explains how to edit, copy, delete, and create media pools.

Creating a media pool

Perform the following steps to create a new media pool.

Before you begin

Perform either of the following: l

If the Match Bar Code Labels attribute is not used for the Library resource, create a label template for the media pool.

Determine a preconfigured label template to use for the media pool.

Procedure

1. In the Administration window, click Media.

2. In the left pane, select Media Pools.

3. From the File menu, select New.

4. In the Name attribute, type a name for the media pool.

A media pool is associated with a label template. Use a name that clearly associates the media pool with the corresponding label template.

5. In the Comment attribute, type a description of the media pool.

84 EMC NetWorker 9.1 Administration Guide x x x x x x x

Levels attribute

Backup Storage

6. Leave the Enabled attribute selected.

7. For the Pool Type attribute, select the media pool type.

l l l l

Backup—Select this type to configure the pool to receive backup data.

Backup clone—Select this option to configure the pool to receive a clone copy of backup data.

Archive—Select this type to configure the pool to receive archive data.

Archive clone—Select this option to configure the pool to receive a clone copy of archive data.

8. In the Label Template attribute, select the matching label template.

9. In the Data Source attribute, select the backup groups that are eligible to back up to this media pool.

10. (Optional), on the Selection Criteria tab, configure the following options: l l l

Devices—Select the devices on which NetWorker can mount volumes for this pool.

Media type required—Select which device type NetWorker can use to label volumes for this pool. You cannot use this attribute when you select an option in the Media type preferred attribute.

Media type preferred—Select the device type that NetWorker should use first to label a volume for this pool. You cannot use this attribute when you select an option in the Media type required attribute.

Note

When you do not configure the Media type required or Media type preferred attribute, you can write data across several volumes of different media types

(for example, magnetic disk and tapes), if the volumes mounted on the storage devices have the appropriate label associated with the media pool.

11. On Configuration tab, configure the following options:

Attribute

Auto Media

Verify

Definition

Select this attribute to perform automated media verification while data is written to a volume labeled for this media pool.

Auto media verification

provides more information.

Max parallelism Increase the value to define the maximum number of simultaneous save streams that NetWorker writes to each device in the pool. The default value for this attribute is 0, which means that the attribute has no effect on other parallelism settings. When you set the Max parallelism attribute to 1, a prolonged delay might occur between the backup of save sets. To resolve this issue, increase the Max

parallelism attribute for the pool resource. However, when you increase the pool parallelism value, the time to recover data on the volume increases.

Using media pools 85

Backup Storage

Attribute Definition

Note

For AFTD and DD Boost devices, the Max nsrmmd count attribute value for a device affects the Max parallelism attribute. For example, consider an AFTD device (AFTD_1) that has a Max sessions attribute value of 20 and a Max

nsrmmd value of 4. Now suppose a backup pool with a Pool parallelism attribute of 1 selects AFTD_1 . The total number of save sessions that NetWorker can start for AFTD_1 is 4, one for each nsrmmd process. Tape and FTD devices can only spawn one nsrmmd process at a time, so if the previous example used a tape device, then the total number of save sessions would be 1.

Recycle from other pools

Recycle to other pools

Select this option to enable NetWorker to use expired volumes that are labeled for other media pools in this pool that have the Recycle to other pools attribute enabled, when the NetWorker server does not have access to blank volumes or volumes eligible for reuse and assigned to this pool.

Select this option to enable NetWorker to use expired volumes that are labeled for this media pool in other pools that have the Recycle from other pools attribute enabled, when the NetWorker server does not have access to blank volumes or volumes eligible for reuse and assigned to the other pool.

Recycle start Defines the time to start the automatic relabel process each day. By default this attribute is empty and the automatic relabeling of recyclable volumes is not done. Use the format

HH:MM.

Automatically relabeling volumes in a media pool

provides more information.

Recycle interval

Max volumes to recycle

Defines the interval between two starts of the automatic relabel processes. The default value is 24:00. Use the format

HH:MM.

Defines the maximum number of recyclable volumes that

NetWorker can relabel during each automatic relabel process. The default value is 200.

Recycle start now

Store index entries

Worm pool/

Create

DLTWORM

Select this attribute to start the automatic relabel process of recyclable volumes for this pool immediately after you create the pool. The default value is No.

For archive pools only. Select this attribute to configure an archive pool that creates client file index entries for the archive save sets. Clear this option to configure an archive pool that will not create client file index entries for the archive save sets.

Supported WORM and DLTWORM tape drives provides more

information about how to create Worm pools.

EMC NetWorker 9.1 Administration Guide

Backup Storage

12. Optionally, on the Restricted Data Zones tab, from the restricted datazone list, select the restricted datazone in which to add the pool.

13. Click OK

If any of the settings for a new media pool match an existing media pool, this message appears:

Pool(s) pool_name has overlapping selection criteria.

If this message appears, review the media pool configuration and modify any overlapping criteria.

14. If you did not select a label template when you create the media pool, a message appears that tells you that NetWorker creates a label template for the media pool, click OK.

Auto media verification

If the Auto Media Verify attribute is enabled, the NetWorker server verifies data written to tape volumes from this media pool. This attribute does not apply to AFTD, file type and Data Domain devices.

Data is verified by repositioning the tape volume to read a portion of the data previously written to the media. The data read is compared to the original data written. This feature does not verify the entire length of the tape.

If the data read matches the data written, verification succeeds.

Media is verified when the following occurs: l l

A volume becomes full while saving and it becomes necessary to continue on to another volume.

A volume goes idle because all save sets being written to the volume are complete.

When a volume fails verification, it is marked full so that the server will not select that volume for future saves. The volume remains full until it is recycled or a user marks it not full. If a volume fails verification while the server is attempting to switch volumes, all save sets writing to the volume are terminated.

Auto media verification should not be used to verify the integrity of the data written to the entire tape. To fully verify the data written to the tape, either restore the tape contents or clone the data.

Automatically relabeling volumes in a media pool

Automatically relabeling a recyclable volume provides the following benefits: l l

You can relabel volumes outside of the backup window without the need for a scripted solution.

NetWorker has access to appendable volumes at the time of a backup or clone, which results in faster backup and clone completion times.

Eligible volumes will not be relabeled if the volume is loaded in a device that is: l l l l

Disabled

In use by an nsrmmd process (for example, during a restore operation)

In read-only mode

Busy

Using media pools 87

Backup Storage

When NetWorker automatically relabels a volume, message to the following appears in the daemon.raw file on the NetWorker server:

"num_of_volumes volumes will be recycled for pool pool_name in jukebox jukebox_name."

Supported WORM and DLTWORM tape drives

NetWorker supports write-once, read-many (WORM) tape drives and media. It is able to recognize the WORM abilities of tape drives and the presence of WORM media in those drives. It also supports the creation of DLTWORM (formerly DLTIce) tapes in drives that are DLTWORM capable.

The following table describes the WORM devices that are supported by the

NetWorker software. For a complete listing of supported devices, refer to the EMC

NetWorker Hardware Compatibility Guide .

Table 23 WORM supported devices

Device Description

HP LTO Ultrium 3 and higher Unique to HP Ultrium-3 and higher: l l

Inquiry VPD page 0xb0, byte 4 bit 0 indicates WORM capable

Read attribute # 0x0408 bit 7 to indicate WORM media present

Quantum SDLT600, DLT-S4, and DLT-V4 (SCSI and SATA)

Any drive with product inquiry data of “*DLT*” tape drive that reports WORM capability the way these drives do

(“Quantum” not required in the vendor inquiry data): l l

Inquiry data VPD page 0xc0, byte 2, bit 0 to indicate

WORM capable

Read attribute # 0x0408 bit 7 to indicate WORM media present

Sony AIT-2, AIT-3, AIT-4, and

SAIT

Any drive with “Sony” in the vendor inquiry data that reports

WORM capability like these drives do: l l

Mode sense page 0x31, byte 5 bit 0 indicates WORM capable

Mode sense byte 4 bit 6 indicates WORM tape present

IBM 3592

STK 9840A/B/C, 9940B,

T10000

Unique to IBM 03592: l l

Mode sense page 0x24, byte 7 bit 4 indicates WORM capable

Mode sense page 0x23, byte 20 bit 4 indicates WORM tape present

Any drive with STK as the vendor data that reports WORM capability like these: l l

Standard inquiry data byte 55 bit 2 indicates WORM capable

Request sense data byte 24 bit 1 indicates WORM tape present

EMC NetWorker 9.1 Administration Guide

Backup Storage

Table 23 WORM supported devices (continued)

Device

IBM LTO Ultrium 3 and higher, and Quantum LTO

Ultrium 3 and higher

Description

These drives use the SCSI-3 method to report WORM capabilities, so there is not a match against any of the inquiry data. Any drive that does not match the inquiry data patterns listed above will have the SCSI-3 method applied to them: l l l

Inquiry data VPD page

0xb0, byte 4, bit 0 indicates WORM capable

Mode sense page

0x1d, byte 2 bit 0 indicates WORM tape present

Byte 4, bits 0,1: label restrictions include

- 00 indicates no overwriting allowed

- 01 indicates some labels can be overwritten

Byte 5, bits 0,1: filemark overwrite restrictions

- 0x02: any filemark at EOD can be overwritten except for the one closest to the beginning of the tape

- 0x03: any filemark at EOD can be overwritten

The WORM and DLTWORM attributes determine whether or not the NetWorker software will back up to a write once-read many (WORM) tape. You can apply these tape attributes to any pool.

Note

Various Quantum drive models (SDLT600, DLT-S4, and DLT-V4) have the ability to create WORM tapes from ordinary blank DLT tapes supported by that particular drive.

You cannot recycle an existing NetWorker tape to create a DLTWORM volume without first having bulk-erased the tape. When the DLTWORM attribute is set, labeling one of these drives into a WORM pool causes the Quantum drive to make the current tape a WORM tape.

Savegroups that belong to pools that have either the WORM or DLTWORM attribute set, are considered to be WORM savegroups.

How to identify WORM media

Since WORM media cannot be reused, the tapes are uniquely identified as such so that they are only used when required. As shown in this figure, a (W) is appended to

Using media pools 89

Backup Storage the volume names displayed in the Volumes window. If a volume is both read-only and

WORM, an (R) is appended to the volume name.

Figure 8 Identifying WORM tapes in the NetWorker Console

Note

Since WORM tapes can only be used once, attempting to relabel a WORM tape always results in a write protection error. With the exception of pool selection and relabeling, the NetWorker software treats WORM tapes exactly the same as all other types of tape.

Determining WORM and DLTWORM capability

Procedure

1. In the Administration window, click Devices.

2. Select the drive, right-click, and select Properties.

3. Click the Information tab and observe the WORM capable and DLTWORM capable attribute settings. NetWorker automatically sets these attributes and, consequently, they are read-only and cannot be changed.

Note

The WORM capable and DLTWORM capable attributes are dimmed out when the device in use is WORM capable but does not support DLTWORM (not a

Quantum DTL-type drive).

Configuring WORM and DLTWORM support

The following table describes WORM and DLTWORM attributes.

Table 24 WORM/DLTWORM attributes

Attribute

WORM pools only hold WORM tape

Description

By default, the NetWorker software only allows WORM tapes into WORM pools.

Deselecting this option lets you add new

(non-WORM) tapes to a WORM pool. This is

90 EMC NetWorker 9.1 Administration Guide

Backup Storage

Table 24 WORM/DLTWORM attributes (continued)

Attribute

WORM tapes only in WORM pools

WORM capable

DLTWORM capable

WORM pool create DLTWORM

Description

useful when you need WORM functionality but do not have WORM tapes available.

By default, NetWorker only lets you label

WORM tapes into WORM pools. Clear this option when:

You do not want to segregate WORM tapes within WORM pools.

A volume is needed to complete a group and a non-WORM tape is unavailable.

This attribute indicates that this drive supports the use of WORM media.

This attribute indicates that this drive can create DLTWORM tapes from a blank tape.

This pool should hold WORM tapes

(depending on the setting of “WORM pools only hold WORM tape” in the server).

If selected, before the NetWorker software labels a tape in a drive capable of creating

DLTWORM volumes, NetWorker will try to convert the tape into a DLTWORM tape. If that conversion fails, the labeling for that tape will fail. If a tape drive in a pool where this attribute is set cannot create DLTWORM tapes, (that is, the tape drive is not a

Quantum SDLT600, DLT-S4 or DLT-V4 tape drive, this attribute is simply ignored.

Refer to the Quantum web site for information on which tapes can be converted to DLTWORM tapes. Not all firmware revisions for all of these devices support WORM operation. Check the tape drives website to make sure that your drive has up-to-date firmware.

Procedure

1. In the Administration window, click Media.

2. In the left pane, select Media Pools.

3. In the right pane, select the appropriate pool.

4. Right-click and select Properties.

5. Click the Configuration tab and select one of these WORM tape handling attributes: l

WORM pools only hold WORM tapes

Using media pools 91

Backup Storage

92 l

WORM tapes only in WORM pools

6. Click OK when finished making the necessary selections.

Note

If you attempt to assign a non-WORM capable drive to a WORM pool an error message is generated.

Editing a media pool

Perform these steps to edit an existing media pool.

Note

You cannot change the name of a media pool. Preconfigured media pools cannot be modified.

Procedure

1. In the Administration window, click Media.

2. In the left pane, select Media Pools.

3. In the right pane, perform one of the following tasks: l l

To modify multiple attributes in a single configuration resource by using the

Media Pool Properties window, right-click the staging configuration and select Properties.

Comment cell and select Edit Comment.

Note

4. Make any required changes, then click OK.

Copying a media pool

Perform these steps to create a copy of a pool resource.

Procedure

1. In the Administration window, click Media.

2. In the left pane, select Media Pools.

3. In the right pane, select the media pool.

4. From the Edit menu, select Copy. The Create Media Pool dialog box appears, containing the same information as the media pool that was copied, except for the Name attribute.

5. In the Name attribute, type a name for the new media pool.

6. Edit any other attributes as appropriate, and click OK.

EMC NetWorker 9.1 Administration Guide

Backup Storage

Deleting a media pool

You can delete a media pool only if the media database does not contain information about active volumes that are labeled for the media pool. You cannot delete a preconfigured media pool.

Procedure

1. In the Administration window, click Media.

2. In the left pane, select Media Pools.

3. In the right pane, select the media pool.

4. From the File menu, select Delete.

5. When prompted, click Yes to confirm the deletion.

Storage nodes

Storage nodes (including the NetWorker server) are host computers with attached storage devices. A storage node has the physical connection and ownership of the attached devices, but the NetWorker server maintains the client file index and media database. With the NetWorker software, client data can be routed directly to a storage node’s storage devices without the data first going to the NetWorker server.

A storage node may be a client of the NetWorker server, although this is not a requirement. However, the storage node must have the NetWorker client software installed.

From the NetWorker server, typical storage tasks can be performed, such as: l

Mounting and labeling volumes for the storage node devices.

Configuring NetWorker resources associated with the storage nodes.

Only users who have the Configure NetWorker privilege can add to or change the configuration of the NetWorker server, media devices, and libraries. The EMC

NetWorker Security Configuration Guide provides more information.

Requirements

To operate the NetWorker software with storage nodes, certain requirements must be met.

On UNIX systems, this software must be installed on the storage nodes. The packages must be installed in the following order: l

1. NetWorker client software

2. NetWorker storage node software

On Windows systems, the Storage Node Option must be installed. The Storage

Node Option installs both the NetWorker client and storage node software.

Licensing

The EMC NetWorker Licensing Guide provides information on NetWorker licensing support for storage nodes.

Storage nodes 93

Backup Storage

Storage node configuration

The following sections provide the procedures for configuring a NetWorker storage node.

Configuring the Linux host as a storage node

Configure a storage node host to manage the data protection activities on a host that is not the NetWorker server.

Note

Do not use a storage node on the CloudBoost appliance and do not create devices on the storage node for CloudBoost.

Procedure

1. Ensure that the storage node software and required enabler codes have been installed on the host.

2. In the NetWorker server Administration interface, click the Devices view.

3. From the navigation tree, right-click Storage Nodes, and select New.

The Create Storage Node window appears, with the General tab displayed.

4. Set the Identity attributes: a. In Name, specify the hostname of the NetWorker storage node.

b. In Type of Storage Node, select SCSI.

5. In the Status attributes, review or set the storage node status: a. Storage node is configured indicates whether a device has already been configured on this storage node.

b. Enabled indicates whether the storage node is available for use: l l

Yes indicates available state.

No indicates service or disabled state. New device operations cannot begin and existing device operations may be canceled.

c. Ready indicates whether the storage node is ready to accept device operations.

6. Set the Device Management attributes: a. In Max active devices, set the maximum number of devices that NetWorker may use from this storage node in a DDS environment.

b. In AFTD allowed directories, for AFTD devices, type the pathnames of directories on the storage host where AFTDs are allowed to be created.

c. In mmds for disabled devices, select a nsrmmd (data mover) option: l l

To start nsrmmd processes for disabled devices, select Yes.

To not start nsrmmd processes for disabled devices, select No.

d. In Dynamic nsrmmds, for AFTD or DD Boost devices, select whether nsrmmd processes on the storage node devices are started dynamically.

94 EMC NetWorker 9.1 Administration Guide

Backup Storage l l

Selected (Dynamic mode): NetWorker starts one nsrmmd process per device and adds more only on demand, for example, when a device's

Target sessions is reached.

Unselected (Static mode): NetWorker runs all available nsrmmd processes.

In environments where unattended firewall ports must be restricted for security reasons, the storage node settings for mmds for disabled devices and Dynamic nsrmmds unselected (static mode) offer more control because they cause all available nsrmmd firewall ports to be attended by running nsrmmd processes.

7. Select the Configuration tab.

8. In Scanning, set the attributes for SCSI library target devices on this storage node: a. In Device Sharing Mode, select an option: l l l

Server Default uses the NetWorker server setting for device sharing.

Maximal Sharing allows sharing of all devices.

No Sharing disables device sharing.

b. In Search all LUNs, select an option: l l

For NetWorker to detect all LUNs (Iogical unit numbers), select Yes.

Detection can be time consuming.

For NetWorker to stop searching at the first available LUN, select No

(the default).

c. In Use persistent names, choose whether NetWorker uses persistent device names specific to the storage host operating system when performing device discovery and autoconfiguration operations.

d. In Skip SCSI targets field: l l l

If the storage node type is set to SCSI, list any SCSI targets to exclude from backup operations, one per line.

The format is bus.target.lun where the target and LUN fields are optional.

You can exclude a maximum of 63 targets.

9. For AFTD or DD Boost devices, configure the following settings in Advanced

Devices: l l

In Server network interface, type the unique network interface hostname of the NetWorker server to be used by the storage nodes.

In Clone storage nodes, list by priority the hostnames of the storage nodes to be used for the save or “write source” side of clone operations originating from this storage node as the “read source.” The clone operation selects the first storage node in this list that has an enabled device and a functional nsrmmd process.

n n

If the Clone storage nodes attribute does not contain a value, then the device operations use the value that is defined in the Clone storage

nodes attribute for the Storage Node resource that was created for the

NetWorker server.

If the Clone storage nodes attribute for the storage node resource is empty, then device operations use the values that are defined in Storage

Storage node configuration 95

Backup Storage

nodes attribute for the client resource that was created for the

NetWorker server.

In backup-to-disk environments, it is possible for a single backup volume to be shared by multiple storage devices on different storage nodes. This can result in an ambiguous clone write source.

10. Click OK.

Modifying the timeout attribute for storage node operations

An attribute named nsrmmd Control Timeout, which is set during NetWorker server configuration, configures the amount of time a NetWorker server waits for a storage node request to be completed. If the timeout value is reached without completion of the request, the operation stops and an error message is logged. The default value assigned to Nsrmmd Control Timeout is five minutes.

Procedure

1. In the server’s Administration interface, click the Configuration button.

2. Select View

Diagnostic Node.

3. Right-click the NetWorker server in the left pane and select Properties.

4. Select the Media tab.

5. Modify the attributes as appropriate and click OK.

Configuring timeouts for storage node remote devices

Timeouts that determine how long to wait for mount requests on a storage node remote device before the save is redirected to another storage node are set in the

Properties window of a device.

The Storage Node Devices area of the tab includes these attributes related to storage node timeouts: l

Save Mount Timeout l

Save Lockout

Save Mount Timeout and Save Lockout attributes change the timeout of a save mount request on a remote device.

If the mount request is not satisfied within the time frame specified by the Save

Mount Timeout attribute, the storage node is locked out from receiving saved data for the time specified by the Save Lockout attribute.

The default value for Save Mount Timeout is 30 minutes. The default value for Save

Lockout is zero, which means the device in the storage node continues to receive mount requests for the saved data.

Note

The Save Mount Timeout applies only to the initial volume of a save request.

To modify the Save Mount Timeout and Save Lockout attributes, perform the following steps.

Procedure

1. In the server’s Administration interface, click the Devices button.

2. Select View

Diagnostic Node.

3. Right-click the remote device and select Properties.

96 EMC NetWorker 9.1 Administration Guide

Backup Storage

4. Select the Advanced tab.

5. Modify the attributes as appropriate and click OK.

Balancing the load on the storage node

The Save Session Distribution feature allows you to configure how NetWorker distributes save sessions between the storage nodes.

Note

This feature is not available for clone and recover operations.

You can apply this feature to all NetWorker clients or to selected clients. This feature has two options: l l

Max sessions—Distributes save sessions that are based on the setting in the Max sessions attribute in storage node device resource. This is the default distribution method.

Target sessions—Distributes save sessions that are based on the setting defined in the Target session attribute in each storage node device resource.

The Target sessions option is more likely to spread the backup across multiple storage nodes. The Max sessions option is more likely to concentrate the backup load on fewer storage nodes.

When you select the Max sessions option, the NetWorker server distributes the save sessions for a client among eligible storage nodes as follows:

1. Identifies the available storage nodes in the NetWorker client’s storage node affinity list.

2. Uses an available device on the first storage node in the list that is working below its Target sessions level.

3. When all devices on the first storage node are running at their target sessions level but some are running below their max sessions level, then NetWorker uses the least loaded device.

4. Continues until all available devices on all storage nodes in the client’s storage node affinity list are in use.

When you select the Target sessions option, the NetWorker server distributes save sessions among eligible storage nodes as follows:

1. Identifies the available storage nodes in the storage node affinity list for the client.

2. Uses an available device on the first storage node in the list that is working below its Target sessions level.

3. When all devices on the first storage node are running at their target sessions levels, continue to the next storage node even if some devices are running below their max sessions level.

4. When all devices on all eligible storage nodes are running at their target sessions level, use the least loaded device that is running below its max session value.

5. Continues to send data to the least loaded device that is running below the max session value, until all devices on all available storage nodes are running at their max session levels.

Be aware of the following performance considerations for storage node load balancing: l

Depending on the configuration of the backup environment, there is a potential to shorten the backup times by using the device Target session option rather than the device Max session option. However, using the device Target sessions option

Storage node configuration 97

Backup Storage

98 l with the checkpoint restart feature can result in slower recovery times because a single save set is more likely to be spread across multiple storage nodes.

Each NetWorker client has a storage node affinity list. The Save sessions distribution feature can only distribute a backup session for a client to multiple storage nodes when the client resource has two or more storage nodes in its storage node affinity list. The storage node affinity list is specified on the Globals

(2 of 2) tab in the NetWorker Client Properties window.

Configuring the storage node affinity list for a client

Storage node affinity is a feature that determines which NetWorker servers and storage nodes receive the data from a client. Define the storage node affinity list in the Storage Nodes attribute of the Client resource.

For most Client resources, the default setting for the Storage Nodes attribute is nsrserverhost, which represents NetWorker server host. To configure the NetWorker server to direct the data for a client to a storage node device, modify the Storage

Nodes attribute and specify the name of the storage node in the Storage Nodes attribute of the Client resource on a line above the default nsrserverhost entry.

If you create the Client resource for a storage node after you create the remote device on the storage node, the default setting of the Storage Nodes attribute is the storage node and the NetWorker server.

To modify the Storage Nodes attribute for a client, perform the following steps:

Procedure

1. In the Administration window, click Protection.

2. In left navigation pane, expand Clients, right-click the appropriate client, and select Properties.

3. On the Globals (2 of 2) tab in the Storage Nodes attribute, specify the hostname of the storage node, and then click OK.

Results

The NetWorker software directs the client data to the first storage node in the affinity list with an enabled device, capable of receiving the data. The NetWorker software sends additional saves to the next storage node in the storage node affinity list that is based on criteria that are specified in

Balancing the load on the storage node

on page

97.

Specifying storage node load balancing

By default, NetWorker balances client backups across storage nodes that are based on the Max sessions attribute for each device on the storage node. If you choose to balance storage node loads by Max sessions, you can override this setting for selected clients.

Procedure

1. On the Administration window, click the Server button.

2. From the View menu, select Diagnostic Mode.

3. Right-click the NetWorker server in the left pane, and select Properties.

4. On the General tab, select a value from the Save session distribution list: l

If you select Target sessions, then the NetWorker server balances the backups for all NetWorker clients across the storage nodes, based on device target session value. The NetWorker server ignores the value that is defined in Save session distribution attribute for each NetWorker client.

EMC NetWorker 9.1 Administration Guide

Backup Storage l

If you select Max sessions, then you can still override this value for selected

NetWorker client resources by setting the Save session distribution attribute in the client resource.

5. Click OK.

Overriding the save session distribution method for selected clients

If you selected Max sessions as the Save session distribution method for the

NetWorker server, you can use the following procedure to override the setting for selected clients.

Procedure

1. On the Administration window, click the Protection button.

2. In the left navigation pane, expand Clients.

3. Right-click the appropriate client and select Properties.

4. On the Globals (1 of 2) tab, select Target sessions from the Save session

distribution list.

5. Click OK.

Performance considerations for storage node load balancing

Be aware of the following performance considerations for storage node load balancing: l l

Depending on how your backup environment is configured, there is a potential to shorten backup times by using the device target session option rather than the device maximum session option. However, using the device target sessions option with the Checkpoint restart feature can result in slower recovery times because a single save set is more likely to be spread across multiple storage nodes.

Each NetWorker client has a storage node affinity list. The Save sessions distribution feature can only distribute a NetWorker client’s backup sessions to multiple Storage nodes if the client has two or more storage nodes in its storage node affinity list. The storage node affinity list is specified on the Globals (2 of 2) tab in the NetWorker Client Properties window.

Multiplexing

Multiplexing is the ability to write multiple data streams simultaneously to the same storage device. It is often more efficient for the NetWorker server to multiplex multiple save sets to the same device. There are also times when limiting the number of data streams to a particular device improves performance of the NetWorker environment.

Use the Target sessions, Max sessions, and Pool parallelism attributes to increase or limit the number of data streams that NetWorker writes to a device.

Target sessions

Use the Target sessions attribute on the Configuration tab of the Device resource to define the optimal number of backup sessions to assign to an active device.

Target sessions is not a hard limit; to set a hard limit for the number of sessions that a particular device can accept, use the Max sessions attribute.

The Target sessions attribute aids in load balancing devices by determining when the

NetWorker software should write save streams to a device.

When a save session starts, the following actions occur: l

If a device is already receiving the number of backup sessions determined by the target sessions value, the NetWorker server uses the next underutilized device for the backups.

Storage node configuration 99

Backup Storage l

If all available devices are receiving the number of backup sessions determined by their target sessions value, the NetWorker server overrides the set value and uses the device with the least activity for the next backup session.

Because it is often more efficient for the NetWorker server to multiplex multiple save sets to the same device, rather than write each save set to a separate device, the

NetWorker server attempts to assign to each device a number of save sets, up to the value of target sessions, before assigning a save set to another device.

NOTICE

When the NetWorker software assesses how many devices need to be involved in multiple save streams assignments with the same storage node, the device with the lowest target session value is used as a reference.

Max sessions

The Max sessions attribute on the Configuration tab of the Device resource defines the maximum number of save sessions for a device. The max sessions value is never less than the target sessions value.

Bootstrap backup on a storage node

When the NetWorker server backup action performs a backup of the bootstrap save set, the data writes to a device that is local to the NetWorker server. You cannot back up bootstrap data to a remote device, but you can clone or stage the bootstrap to a remote device. When you recover a bootstrap save set, you must recover the data from a local device.

Staging bootstrap backups

You can direct bootstrap backups to a disk device such as an AFTD or FTD device.

However, if you stage a bootstrap backup to a volume on another device, NetWorker reports the staging operation as complete although the “recover space” operation has not started, and the bootstrap remains on the original device. Therefore, if the staged bootstrap is accidentally deleted, you can recover the bootstrap from the original disk.

Was the bootstrap staged?

on page 579 in the "Recovery" chapter describes how to

recover a bootstrap from the original disk.

Also, if the bootstrap data is not staged from the original disk, the data on the original disk is subject to the same retention policies as any other save set backup and is, therefore, deleted after the retention policy has expired.

Troubleshooting storage node affinity issues

If a backup fails because of a problem related to the storage node affinity, a message similar to the following might appear:

100 no matching devices; check storage nodes, devices or pools

Possible causes for this error message include: l l l

No enabled devices are on the storage nodes.

The devices do not have volumes that match the pool required by the backup request.

All devices are set to read-only or are disabled.

For example, if the client has only one storage node in its Storage Node list, and all devices on that storage node are disabled, fix the problem and then restart the backup.

EMC NetWorker 9.1 Administration Guide

Backup Storage

Complete one of the following actions to fix the problem: l l l l

Enable devices on one of the storage nodes in the storage node list for the client.

Correct the pool restrictions for the devices in the storage node list.

Configure an additional storage node that has enabled devices that meet the pool restrictions.

Set one of the devices to read/write.

Configuring a dedicated storage node

All devices created on storage nodes, except the devices for the NetWorker server include the Dedicated Storage Node attribute. A dedicated storage node can only back up data that originates from the storage node host. When you configure a storage node as a dedicated storage node, you require a Dedicated Storage Node license.

After you create a storage node, perform the following steps to configure the storage node as dedicated.

Procedure

1. On the Administration window, click Devices.

2. In the left navigation pane, expand Storage Nodes, right-click the storage node, and then select Properties.

3. On the Configuration tab, in the Dedicated Storage Node option, select Yes.

4. Click OK.

Troubleshooting storage nodes

This section provides troubleshooting information about storage nodes.

Storage node affinity errors

A storage node affinity problem may exist when a backup fails with an error message similar to the following:

No matching devices; check storage nodes, devices or pools

This error message can appear for the following reasons: l l l

All the devices in the storage node are disabled.

Each device in the storage node contains a volume that does not match the pool that the backup request requires.

All the devices in the storage node are set to read-only.

To resolve this error: l l l l l

Enable devices on one of the storage nodes.

Correct the pool restrictions for the devices that are listed in the Storage Nodes attribute of the Pool resource.

Add another storage node that has enabled devices and meets the pool restrictions to the Storage Nodes attribute of the Pool resource.

Write-enable one of the devices.

Adjust the Save Mount Timeout and Save Lockout attributes for in the Device resource for the storage node.

Configuring a dedicated storage node 101

Backup Storage

Storage node timeout errors

If the nsrd process starts on the NetWorker server and detects that a setting for the

NSR_MMDCONTROL variable exists, a message similar to the following appears:

NSR_MMDCONTROL env variable is being ignored use nsrmmd control timeout attribute instead

If you receive this message, perform the following steps.

1. Shut down the NetWorker services.

2. Remove the environment variable setting for

NSR_MMDCONTROL.

3. Restart the NetWorker services.

4. Use NMC to connect to the NetWorker server.

5. Adjust the value of the nsrmmd Control Timeout attribute in the Storage Node resource to the value that was assigned to the NSR_MMDCONTROL variable, or

to a value that best meets the current requirements. Modifying the timeout attribute for storage node operations on page 96 provides more information.

Disk storage devices

NetWorker software supports a variety of different backup to disk (B2D) methods.

These methods all use disk files that NetWorker creates and manages as storage devices. These devices can reside on a computer’s local disk or a network-attached disk. NetWorker supports FTD, AFTD, DD Boost, and cloud disk device types. This section does not cover disk-based devices that emulate other device types, such as virtual tape libraries (VTLs). The EMC NetWorker Licensing Guide provides information about NetWorker B2D and DD Boost licensing.

FTD

A file type device (FTD) is a basic disk device type that has been available for many years. FTDs have limited use and support and this chapter describes them for legacy purposes only.

AFTD

Advanced file type devices (AFTDs) support concurrent backup and restore operations and require the NetWorker DiskBackup Option (DBO) license. AFTDs are supported for the following configurations: l l

A local disk on a NetWorker storage node.

A network-attached disk device that is NFS-mountable to a NetWorker storage node running a Linux or UNIX operating system.

A network-attached disk device that is CIFS-mountable to a NetWorker storage node running on Windows.

The Client Direct feature enables NetWorker clients to back up directly to AFTDs over a CIFS or NFS network, bypassing the storage node. For Client Direct backups, the storage node manages the devices but does not handle the backup data unless the

Client Direct workflow is not available.

DD Boost devices

DD Boost devices reside on Data Domain storage systems that have the DD Boost features enabled. These devices are similar to AFTDs except they store backup data in a highly compressed and deduplicated format. The DD Boost API accesses the DD

Boost devices over a network. NetWorker can perform DD Boost backups through either the NetWorker storage node workflow or the Client Direct file access workflow.

102 EMC NetWorker 9.1 Administration Guide

Backup Storage

The Client Direct workflow enables NetWorker clients with distributed segment processing (DSP) and network access to deduplicate their own backup data and send the data directly to the DD Boost devices. This method bypasses the storage node and frees up network bandwidth. The storage node manages the devices but does not handle the backup data workflow if the Client Direct workflow is available.

If Client Direct backup is not available, NetWorker automatically routes the backup through the storage node where it is deduplicated and sent to the DD Boost devices for storage. Restore operations work similarly. If Client Direct is not available for a restore, then NetWorker performs a traditional storage node recovery.

This guide does not cover DD Boost operations. The EMC NetWorker Data Domain

Boost Integration Guide provides details on DD Boost devices

Cloud devices

Cloud devices are specific to cloud storage services, such as ATMOS. NetWorker accesses cloud services through a private network.

Example environment

The following figure shows various backup-to-disk options deployed in a mixed operating system environment.

Linux/UNIX Storage Node A writes its backups to either of the following: l l n

The AFTD through an NFS connection to Disk Device 1.

The AFTD on Local Disk 1.

Windows Storage Node B uses a CIFS connection to back up to the NAS AFTD on

Disk Device 2.

Data Domain system C writes its backups to a DD Boost device on Local Disk 2.

Figure 9 Example NetWorker disk backup configuration in a mixed backup environment.

Considerations for Client Direct clients

Client Direct backups enable clients to bypass the storage node and back up directly to storage devices. The storage node manages the devices but does not handle the backup data. Device configuration for Client Direct clients depends on the type of storage device and how it is connected to the storage nodes.

A Client Direct backup reduces bandwidth usage and bottlenecks at the storage node, and provides highly efficient backup data transmission.

Example environment 103

Backup Storage

If a Client Direct backup is not available, a traditional storage node backup occurs instead.

Requirements for Client Direct backups

Ensure that the environment meets the following requirements to perform Client

Direct backups: l l l l

NetWorker clients on UNIX/Linux or Microsoft Windows can perform non-root and cross-platform Client Direct backups to AFTDs. The AFTD can be managed by either a UNIX/Linux or a Windows storage node, and can be either local or mountable on the storage node.

To perform non-root and cross-platform Client Direct backups to AFTDs, the

NetWorker server and the storage node software must be version 8.1 or later.

If an NFS server provides the AFTD storage for Client Direct backups, then the

NFS server must permit access by using the NFSv3 protocol with AUTH_SYS

(AUTH_UNIX) authentication. The NFS server also must not restrict access to clients by using only privileged ports.

If you enable checkpoint restart for a client, then Client Direct backups are supported only to AFTDs, and not to DD Boost devices. If a client is enabled for checkpoint restart and a Client Direct backup is tried to a DD Boost device, then the backup reverts to a traditional storage node backup instead.

For Client Direct backups to AFTDs, checkpoint restart points are made at least 15 seconds apart. Checkpoints are always made after larger files that require more than 15 seconds to back up.

Archive operations are not currently supported for Client Direct backups.

Configuring Client Direct backups

Procedure

1. Ensure that the clients that perform Client Direct backups have a network connection and a remote network protocol to reach the storage device.

Windows clients can use a CIFS or NFS path, although a CIFS path generally yields better performance. UNIX clients must use an NFS path.

2. Specify the complete path for the destination device in the Device access

information attribute on the General tab of the Device Properties dialog box for the destination device.

Keep in mind the following points when you specify the path: l l l l

If the storage device is directly attached to a Windows storage node, then the storage node uses a different path than the Client Direct clients. If the storage device is not directly attached to any storage node, then the path is the same for all storage nodes and Client Direct clients.

The device access information path should include multiple access paths to cover local and remote use cases.

To specify an NFS path, use the NFS_host:/path format regardless of whether the AFTD is local to the storage node or mountable on the storage node. Non-root UNIX/Linux NetWorker clients require this NFS format for

Client Direct access.

For Windows Client Direct backups, specify a CIFS path instead of an NFS path. A CIFS path generally yields better performance.

104 EMC NetWorker 9.1 Administration Guide

Backup Storage l

If you are setting up an AFTD on a Windows storage node, specify the CIFS path first. For example: l

\\fileserver\aftd1 fileserver:/aftd1

If you are setting up a UNIX/Linux storage node, specify the NFS path first.

For example:

fileserver:/aftd1

\\fileserver\aftd1

The following figure shows an example set of paths for a CIFS AFTD.

Figure 10 Paths for CIFS AFTD

3. If an NFS server provides the AFTD storage for Client Direct backups, then specify the username and password that is required to access the NFS server for the AFTD in the Remote user and Password attributes on the

Configuration tab of the Device Properties dialog box for the device.

4. Ensure that the Client direct attribute is enabled on the General tab of the

Client Properties dialog box for each Client Direct client.

Client Direct backups are enabled by default.

Select View

Diagnostic Mode in the Administration interface to access the

Client direct attribute in the Client Properties dialog box.

Cloud storage devices compared to other disk device types

NetWorker supports Data Domain Cloud Tier (DD Cloud Tier), CloudBoost, and Atmos

Cloud backup devices. Backup, staging, cloning, and recovery to cloud storage devices are similar to those operations that are performed with conventional devices.

However, cloud storage devices also have unique features.

The following table lists the major similarities with other backup device types as well as the unique features of a cloud storage device.

Table 25 A comparison of cloud devices to other device types

Feature

Staging source

Staging destination

Cloning

DD Cloud Tier device

Yes

CloudBoost device

Yes

AFTD device Tape device

Yes

Cloud storage devices compared to other disk device types 105

Backup Storage

Table 25 A comparison of cloud devices to other device types (continued)

Feature DD Cloud Tier device

Yes

CloudBoost device

Yes

AFTD device Tape device

No Yes * Auto mount and unmount

Data transformation engine

(enables encryption and compression on storage node)

Recover directly from device?

Yes

No***

Yes

Yes****

No **

Yes

No **

Yes

* When the tape is controlled by a tape library.

** Encryption and compression can still be enabled through NetWorker client side directives.

*** The recovery operation automatically clones the data to a Data Domain device for recovery.

**** Except for the bootstrap backup, which you must manually clone to another device.

The NetWorker 9.1.x with CloudBoost 2.1 Integration Guide describes how to use

NetWorker with CloudBoost devices.

Differences between FTDs, AFTDs, and DD Boost devices

The following table lists the functional differences between traditional file type devices (FTDs), AFTDs, and DD Boost devices.

The EMC NetWorker Data Domain Boost Integration Guide provides details on DD Boost devices.

Table 26 Differences between disk devices

Function or operation

Create a device

File type device

(FTD)

Advanced file type device (AFTD)

DD Boost device

Device Property

Window l

Select the media type: file.

UNIX/Linux storage node: local or NFS only.

Windows storage node: local path only. CIFS is not supported for

FTDs.

Device Configuration

Wizard

Device Property

Window

Select media type: adv_file.

UNIX/Linux storage node: local or NFS only.

Windows storage node: local or

CIFS using UNC path or using NFS:

Remote user,

Password.

l l

Device

Configuration

Wizard

Device

Property

Window

Select media type:

Data Domain

106 EMC NetWorker 9.1 Administration Guide

Backup Storage

Table 26 Differences between disk devices (continued)

Function or operation

Storage location

File type device

(FTD)

Advanced file type device (AFTD)

Specified in the

Name

attribute.

Specified in the

Device Access

Information

attribute.

DD Boost device

Specified in the

Device

Access

Information

attribute.

Yes.

Concurrent save set operations

Concurrent AFTD recovery operation limitations

on page

123 provides more information about performing concurrent recovery operations from an

AFTD.

Reclaiming or recovering space l

No.

Volume default capacity for devices l

Yes.

The nsrim program removes both aborted and expired save sets, once every

24 hours, by the

Expiration action, at the time defined in the Server backup workflow (if you have set volume recycle to

Auto).

l l

Aborted save sets immediately removed.

The nsrim program removes expired save sets from the media database once every

24 hours, by the

Expiration action, at the time defined in the Server backup workflow (if you have set volume recycle to

Auto). NetWorker removes space on the

AFTD as specified in the Reclaim Space

Interval of the staging policy.

Reclaims only data that is unique, not required by other existing backups.

NetWorker does not immediately remove aborted save sets, but marks them recyclable. A restarted save can be deduplicated.

Otherwise,

NetWorker removes the aborted save set during the next Recover

Space operation.

If the file type device was used before setting the Volume

Default l

Does not apply.

Differences between FTDs, AFTDs, and DD Boost devices 107

Backup Storage

108

Table 26 Differences between disk devices (continued)

Function or operation

File type device

(FTD)

Capacity attribute, the data for that file type device must be staged or cloned to another device.

Advanced file type device (AFTD)

DD Boost device

AFTD Percentage

Capacity

When file system or volume is full

Save set continuation l

Does not apply.

A setting determines the capacity that

NetWorker software should stop writing to an AFTD: spans from

1% to 100%.

Does not apply.

l l

Waiting message is displayed if no writable volume available or until volume becomes available.

Volume marked full and is no longer available for backups until the volume becomes appendable.

l l l l

Message is displayed stating file system requires more space.

The nsrim program invoked to reclaim space for expired save set on AFTD.

Notification is sent by email stating device is full.

Device waits until space become available. The volume is never marked as full.

Backup to a

DD Boost device fails and stops when full.

Yes.

No. Save sets that start on an AFTD must be completed on the same device.

No. Save sets that start on a

DD Boost device must be completed on the same device.

Data format in device

Client Direct backup: the storage node manages the devices for the NetWorker clients, but the clients send their backup data l l

EMC Open

Tape Format

(OTF).

No.

l l

Save stream (uasm) format (uses less space).

Yes.

Clients send their own backup data directly to the l l

Deduplicated

Yes.

Clients use DD

Boost

EMC NetWorker 9.1 Administration Guide

Backup Storage

Table 26 Differences between disk devices (continued)

Function or operation

directly to the devices via network access, bypassing the storage node.

File type device

(FTD)

Advanced file type device (AFTD)

storage devices. If

Client

Direct backup is not available, a traditional storage node backup is performed.

NetWorker archive operations are not supported for

Client Direct backup.

DD Boost device

DSP functionality to deduplicate their own backup data before sending it directly to the storage devices. If

Client Direct backup is not available, a traditional storage node backup is performed.

NetWorker archive operations are not supported for

Client Direct backup.

Device target and max sessions default values and ranges

There are default values and ranges for device target and max sessions in the

NetWorker Administration interface.

The following table lists the default values for target and max sessions values.

Table 27 Default values and ranges for target and max sessions attributes

Device type

AFTD (traditional storage)

AFTD (including

Data Domain

CIFS/NFS)

Data Domain (DD

Boost)

Cloud

NDMP

Default target sessions

Default max sessions

512

Recommende d sessions*

1 - 32

Range

1 - 1024

1 - 10

Any

1 - 32

1 - 1024

1 - 60

1 - 1024

Device target and max sessions default values and ranges 109

Backup Storage

Table 27 Default values and ranges for target and max sessions attributes (continued)

Device type Default target sessions

Default max sessions

Recommende d sessions*

Range

FTD (traditional) 4

VTL/Tape

(traditional)

1 - 16

1 - 1024

VTL/Tape (Data

Domain /

Deduplicated)

4 32 1 - 1 1 - 1024

* The recommended session values are guidelines only and are subject to bandwidth, data type, and device capabilities.

Advanced file type devices

Advanced file type devices (AFTDs) overcome the main restrictions of traditional file type device (FTD) storage. AFTD storage is designed for large disk storage systems that use a volume manager to dynamically extend available disk space if the disk runs out of space during backup.

The EMC NetWorker Online Software Compatibility Matrix provides a list of supported volume managers.

Memory requirements for AFTD backups

The physical memory requirements for a NetWorker storage node and Client Direct client depends on the peak AFTD usage.

The following is the list of physical memory requirements for AFTD: l l l

Allowing for other types of devices and services on a typical storage node, a storage node should have a minimum of 8 GB of RAM to host AFTDs.

AFTD clients require a minimum of 4 GB of RAM at the time of backup to ensure optimum performance for Client Direct backups. Client Direct backups require client access to the AFTDs on either a CIFS or NFS network.

Each AFTD requires an initial 24 MB of RAM on the storage node and Client Direct client. Each AFTD save session requires an additional 24 MB. To run 10 sessions requires 24 + 240 MB. The default max sessions of 60 sessions per AFTD requires

24 + 1440 MB.

Required AFTD DFA device settings for Hyper-V environments

For Hyper-V environments, when creating a NetWorker AFTD DFA device on an NTFS or ReFS volume, Microsoft requires certain settings.

If the NetWorker AFTD DFA device is created on an NTFS volume, virtual hard disk files must be uncompressed and unencrypted. If the NetWorker AFTD DFA device is created on an ReFS volume, virtual hard disk files must not have the integrity bit set.

Create and configure an AFTD

You can create an AFTD by using either the Device Wizard or the device properties window.

110 EMC NetWorker 9.1 Administration Guide

Backup Storage

Creating an AFTD by using the Device Wizard

If you are creating an AFTD to use the client direct feature, see

Considerations for

Client Direct clients on page 103 for information about specifying network path

information when creating the AFTD.

Procedure

1. In the NMC Enterprise view, double-click the NetWorker managed application to launch its window.

2. In the NetWorker Administration window, select the Devices view.

3. Verify that the path to the storage directory that will contain the AFTDs is allowed.

a. In the navigation tree, select Storage Nodes.

b. Right-click the storage node that you will use, and select Properties.

c. In the AFTD allowed directories list, verify or type the path of the storage directory that will contain the AFTDs.

AFTDs can be created and accessed only by these listed paths. If this list is left empty, there are few restrictions as to where a device path can be created.

d. Click OK.

4. In the navigation tree, right-click Devices, and select New Device Wizard.

5. In the Select the Device Type window, select AFTD and click Next.

6. In the Select Storage Node window, specify the path to the storage directory that will contain the AFTDs.

a. In the Storage Node list, select the storage node that you will use.

b. If the directory for the intended AFTDs is on a different storage node or a remote storage system, select Device storage is remote from this Storage

Node and type the Network Path of the remote host directory that will contain the devices.

For example, if the storage node is a Microsoft Windows system and you use a CIFS AFTD on a remote storage system host, this path could be something like the following:

\\dzone1_storhost2.lss.corp.com\share-1

This storage path is not a device. It is the directory location in which the shared devices are to be created.

7. In Browse or Manual, select which option you will use to specify the pathnames of the devices: l l

Browse Storage Node or network path. The next wizard step will prompt you to browse and add the devices.

Manually enter local or remote device paths. Select this to skip the browse step and manually type unique names for the devices you want to add: n

For remote devices, type the device paths relative to the Network Path that you specified for the storage directory. For example: cifsaftd-1 cifsaftd-2

Advanced file type devices 111

Backup Storage

112 n

For local devices, type the absolute paths to these devices. For example:

C:\cifsaftd-1

C:\cifsaftd-2

Multiple devices for a single volume configuration on page 118 provides

details for shared volumes.

8. If the storage host is remote from the storage node, in the Authentication area, type the appropriate Username and Password to access the storage directory.

9. Click Next.

10. If you selected the Browse option in the previous window: a. In the Select the Device Path window, verify that the storage node shows the path of a storage directory.

b. Add devices to the storage directory by clicking New Folder and typing unique device names. For example: cifsaftd-1 cifsaftd-2 c. Select the new devices to add and click Next.

11. In the Configure Device Attributes window, specify the attributes. If you added multiple devices in the previous window, select each device individually and specify its attributes: a. In NetWorker Device Name, type a unique name for the AFTD device.

For example, for a device on the NetWorker server host storage node: aftd-1

If you configure the device on a storage node host that is not the NetWorker server host, it is a “remote device” and this attribute must be specified with rd= and a colon (:) in the following format (for Microsoft Windows): rd=remote_storagenode_hostname:device_name

For example: rd=dzone1_storhost2:aftd-1 b. (Optional) Add a comment in the Comment field.

c. If Client Direct backup will be used, follow the details in Considerations for

Client Direct clients on page 103.

d. In Target Sessions specify the number of sessions that a nsrmmd data mover process on the device will handle before another device on the host will take the additional sessions. Use this setting to balance the sessions among nsrmmd processes.

If another device is not available, then another nsrmmd process on the same device will take the additional sessions.

Typically, set this attribute to a low value. The default value is 4 for AFTDs.

It may not be set to a value greater than 60.

e. In Max Sessions specify the maximum number sessions the device may handle. If no additional devices are available on the host, then another available storage host takes the additional sessions, or retries are tried until sessions become available.

EMC NetWorker 9.1 Administration Guide

Backup Storage

The default value is 32 for AFTDs, which typically provides best performance. It cannot be set to a value greater than 60.

Note

The Max Sessions setting does not apply to concurrent recover sessions.

f. Click Next.

12. In the Label and Mount device window, if you select the Label and Mount option, specify the attributes for: l l

Pool Type.

Pool to use.

13. In the Review the Device Configuration page: a. Review the settings.

b. Click Configure.

14. In the Check results page: a. Review whether the devices were successfully completed or any messages.

b. Click Finish.

c. To go back, click Back to the appropriate wizard step.

Creating an AFTD by using the Properties window (Linux and UNIX)

Procedure

1. Create one directory for each disk (or partition) to be used for an AFTD.

AFTDs require a directory (folder) to be created in the disk file system that the

NetWorker server or storage node recognizes as the device name (and the destination for the data).

NOTICE

Do not use a temporary directory for AFTDs. The data could be overwritten.

2. In the NetWorker Administration window, click the Devices view.

3. Verify that the path to the storage directory that will contain the AFTDs is allowed.

a. In the navigation tree, select Storage Nodes.

b. Right-click the storage node that you will use, and select Properties.

c. In the AFTD allowed directories list, verify or type the path of the storage directory that will contain the AFTDs.

AFTDs can be created and accessed only by these listed paths. If this list is left empty, there are few restrictions as to where a device path can be created.

d. Click OK.

4. In the navigation tree, right-click Devices and select New.

The Create Device window opens, with the General tab selected. The Identity area might show a default device name in the Name field.

Advanced file type devices 113

Backup Storage

114

5. In the Identity area, set the following attributes: a. In the Name attribute, type the name of the directory that you created for the AFTD.

For example: aftd-1

If you configure the device on a separate storage node host that is not the

NetWorker server host, it is a remote device and this Name attribute must be specified with rd= in the following format:

rd=remote_snode_hostname:device_name

For example:

rd=snode-1:aftd-1

b. (Optional) Add a comment in the Comment field.

c. In the Device Access Information attribute, provide complete paths to the device directory. You can provide alternate paths for the storage node and for Client Direct clients, for example:

For non-root or cross-platform Client Direct access:

For non-root or cross-platform Client Direct access to an AFTD, do not specify an automounter path or a mounted path. Instead, specify the path in the host:/path format, even if the AFTD is local to the storage node.

For example:

NFS_host:/path

where: l l

NFS_host is the hostname of the NFS file server path is the NFS-mountable path that is exported by the file server

This format is required to allow Client Direct access for Windows or nonroot UNIX clients.

Note

Non-root Client Direct access to an NFS AFTD is supported only with the

NFSv3 protocol and AUTH_SYS authentication on the NFS host. For

Client Direct access to an AFTD when the backup client is able to run as root on the AFTD host, provide a mount point or automounter path.

Note

For example, for an NFS-mounted device:

/mnt/aftd-1

/net/storho-1/snode-1/aftd-1 where: n n n

aftd-1

is the storage device directory name

storho-1

is the storage system hostname

snode-1

is the storage node hostname

The first path enables the storage node to access the device via its defined mount point. The second path enables Client Direct clients to

EMC NetWorker 9.1 Administration Guide

Backup Storage use the automounter path to directly access the device, bypassing the storage node.

d. In the Media Type field, select adv_file, for the AFTD.

Considerations for Client Direct clients on page 103 provides additional

details for Client Direct configurations.

Multiple devices for a single volume configuration on page 118 provides

additional details for shared volumes.

6. In the Status area, ensure that the Auto Media Management tape feature is not enabled.

7. In the Cleaning area, leave the options for cleaning at their default (disabled) settings, so that automatic cleaning is not invoked.

8. Select the Configuration tab.

9. In the Save Sessions area, set the number of concurrent save sessions

(streams) and the number of nsrmmd (data mover) processes the device may handle: l l l

Target Sessions is the number of sessions that a nsrmmd process on the device will handle before another device on the host will take the additional sessions. Use this setting to balance the sessions among nsrmmd processes.

If another device is not available, then another nsrmmd process on the same device will take the additional sessions.

Typically, set this attribute to a low value. The default values are 4 for

AFTDs and 6 for DD Boost devices. It may not be set to a value greater than

60.

Multiple devices for a single volume configuration

on page 118 provides details on volume sharing.

Max Sessions is the maximum number sessions the device may handle. If no additional devices are available on the host, then another available storage host takes the additional sessions, or retries are attempted until sessions become available.

The default values are 32 for AFTDs and 60 for DD Boost devices, which typically provides best performance. It cannot be set to a value greater than

60.

The Max Sessions setting does not apply to concurrent recover sessions.

Max nsrmmd count limits the number of nsrmmd processes that can run on the device. Use this setting to balance the nsrmmd load among devices. The default value is 4.

To modify this value, first adjust the sessions attributes, apply, and monitor the effects, then update max nsrmmd count.

At least one nsrmmd process is reserved for restore or clone operations.

10. In the Local Backup area, leave Dedicated Storage Node at No (the default).

11. In the Remote Host area, if an NFS path is specified in the Device Access

Information, then type a Remote User name and Password.

The remote username is the name of the user on the NFS server. It is recommended that you also specify the numeric user id (UID) of that user. Do this by appending a colon (:) and the UID after the username, for example,

user_name:4242

Advanced file type devices 115

Backup Storage

116

Note

If the device username is changed after labeling, manual action may be required to change the owner of all files and directories in the AFTD. NetWorker will try to perform this automatically during the next operation, however the ability to do so depends on the security configuration of the file server where the AFTD storage resides.

12. Click OK when the configuration is complete.

13. If a new password for an AFTD is provided, unmount and re-mount the device to ensure that the change takes effect.

Creating an AFTD by using the Properties window (Windows)

You can configure an AFTD on a storage node running Microsoft Windows.

Procedure

1. Create one directory for each disk (or partition) to be used for an AFTD.

AFTDs require a directory (folder) to be created in the disk file system that the

NetWorker server or storage node recognizes as the device name (and the destination for the data).

NOTICE

Do not use a temporary directory for AFTDs. The data could be overwritten.

2. In the NetWorker Administration window, click the Devices view.

3. Verify that the path to the storage directory that will contain the AFTDs is allowed.

a. In the navigation tree, select Storage Nodes.

b. Right-click the storage node that you will use, and select Properties.

c. In the AFTD allowed directories list, verify or type the path of the storage directory that will contain the AFTDs.

AFTDs can be created and accessed only by these listed paths. If this list is left empty, there are few restrictions as to where a device path can be created.

d. Click OK.

4. In the navigation tree, right-click Devices and select New.

The Create Device window opens, with the General tab selected. The Identity area might show a default device name in the Name field.

5. In the Identity area, set the following attributes: a. In the Name attribute, type the name of the directory that you created for the AFTD.

For example: aftd-1

If you configure the device on a separate storage node host that is not the

NetWorker server host, it is a remote device and this Name attribute must be specified with rd= in the following format:

EMC NetWorker 9.1 Administration Guide

Backup Storage

rd=remote_snode_hostname:device_name

For example:

rd=snode-1:aftd-1

b. (Optional) Add a comment in the Comment field.

c. In the Device Access Information attribute, provide complete paths to the device directory. You can provide alternate paths for the storage node and for Client Direct clients, for example: l

For an AFTD on the storage node’s local disk, which it shares via CIFS: l

E:\aftd-1

\\snode-1\aftd-1

The first path enables the storage node to access the device via its local drive. The second path enables Client Direct clients to access the device directly, bypassing the storage node.

For a CIFS-mounted AFTD, specify the complete paths of the directory that is created by using the Universal Naming Convention (UNC), for example:

\\CIFS_host\share-point-name\path d. In the Media Type field, select adv_file, for the AFTD.

Considerations for Client Direct clients on page 103 provides additional

details for Client Direct configurations.

Multiple devices for a single volume configuration on page 118 provides

additional details for shared volumes.

6. In the Status area, ensure that the Auto Media Management tape feature is not enabled.

7. In the Cleaning area, leave the options for cleaning at their default (disabled) settings, so that automatic cleaning is not invoked.

8. Select the Configuration tab.

9. In the Save Sessions area, set the number of concurrent save sessions

(streams) and the number of nsrmmd (data mover) processes the device may handle: l l

If another device is not available, then another nsrmmd process on the same device will take the additional sessions.

Typically, set this attribute to a low value. The default values are 4 for

AFTDs and 6 for DD Boost devices. It may not be set to a value greater than

60.

Multiple devices for a single volume configuration

on page 118 provides details on volume sharing.

Advanced file type devices 117

Backup Storage

118 l

The default values are 32 for AFTDs and 60 for DD Boost devices, which typically provides best performance. It cannot be set to a value greater than

60.

The Max Sessions setting does not apply to concurrent recover sessions.

Max nsrmmd count limits the number of nsrmmd processes that can run on the device. Use this setting to balance the nsrmmd load among devices. The default value is 4.

To modify this value, first adjust the sessions attributes, apply, and monitor the effects, then update max nsrmmd count.

At least one nsrmmd process is reserved for restore or clone operations.

10. In the Local Backup area, leave Dedicated Storage Node at No (the default).

11. In the Remote Host area, if a network path is specified in the Device Access

Information, then type a Remote User name and Password.

12. Click OK when the configuration is complete.

13. If a new password for an AFTD is provided, unmount and re-mount the device to ensure that the change takes effect.

AFTD device target and max sessions

The default settings for AFTD target sessions and max device sessions typically provide optimal values for AFTD performance: l

Device target sessions is 1 l

Device max sessions is 32 to avoid disk thrashing

If required, both device target, and max session attributes can be modified to reflect values appropriate for the environment.

Note

The Max Sessions setting does not apply to concurrent recover sessions.

Multiple devices for a single volume configuration

In some environments, a configuration of multiple devices that share a single

NetWorker storage volume can result in performance gains. For example, a read or write request can be sent to the storage node that is closest to the requestor.

However, for some use cases and environments concurrent read/write operations to a single volume from many storage nodes could result in disk thrashing that impacts performance.

Multiple devices can be created on separate storage nodes or on the same storage node. Each device must be created separately, have a different name, and must correctly specify the path to the storage volume location.

For example, if you create three devices, one on the NetWorker server host named

“dzone1” (that uses the server’s local storage node) and two remote devices (rd) on remote storage nodes, the Name attributes for the three devices, each created separately, might be specified by different aliases as follows: aftd-1a rd=dzone1-sn2:aftd-1b rd=dzone1-sn3:aftd-1c

The Device Access Information for each of these aliases would specify a single directory that must be specified as a valid complete path. For example, if a directory is

EMC NetWorker 9.1 Administration Guide

Backup Storage named “aftd-1” on the storage host named “storho1,” the path might be specified as follows: l l

If the storage node uses an automounter:

/net/storho1/dzone1/aftd-1

If the storage node uses an explicit system mountpoint, you might specify one of the following paths: n n n

/mnt/storho1/dzone1/aftd-1

/mnt/dzone1/aftd-1 storho1:/dzone/aftd-1

AFTD concurrent operations and device formats

The following operations can be performed concurrently on a single storage node with an AFTD: l l l

Multiple backups and multiple recover operations

Multiple backups and one manual clone operation

Multiple backups and one automatic or manual staging operation

It might be required to increase the server parallelism value to complete the concurrent operations with an AFTD device when the number of simultaneous save sessions reaches the maximum value for server parallelism.

For example, if server parallelism is set to 4, and there are 4 simultaneous saves going to an AFTD, set the server parallelism to 5 to complete a concurrent clone/stage operation from this AFTD while the four saves are in progress.

Note

Starting with NetWorker 8.0, multiple clone sessions can be run from a single AFTD or

DD Boost device if each clone is written to a dedicated tape device. However, the number of clone sessions that can be run is limited by the value in the device’s max nsrmmd count attribute.

Create and configure an AFTD on page 110 provides more

information.

Labeling and mounting an AFTD

If there are multiple volumes in the pool, you can select an available volume to associate with the device.

Procedure

1. Right-click the AFTD storage device and select Label.

The Label dialog box appears.

2. In the Pools field, select the media pool to be used for the device.

A label for the storage device is generated and displays in the Volume Label field. The label name is based on the label template for the selected pool.

It is recommended to use a pool dedicated to AFTD backup devices only.

NOTICE

If an existing volume is re-labeled, a warning is issued. The data previously stored on the volume will be lost and this action cannot be undone. Mounting the volume without labeling provides access to previous data.

Advanced file type devices 119

Backup Storage

3. Select Mount after labeling and click OK.

Insufficient AFTD disk space

When an AFTD runs out of disk space, the current backup is interrupted and the following message displays:

Waiting for more available space on filesystem device-name

Immediately following the message, the action that is associated with the "Filesystem

Full — Recover adv_file Space" notification occurs. By default, the action for this notification uses the nsrim command to delete expired save sets. If enough space is cleared, the backup continues. If the recycle setting for the volume is manual, then the expired save sets are not removed from the volume.

The AFTD deletes expired save sets depending on the retention policy and the recycle setting. If sufficient storage space is not available after 10 minutes from when the expired savesets begin deletion, the associated "Filesystem Full—Waiting for adv_file

Space" notification action occurs. By default, an email notification is sent to the root user on the NetWorker server on UNIX and Linux, and a message is logged in the media log file in NetWorker_install_path\logs on Windows.

When the notification is sent, and the message is logged in the media log file, the backup stops until space is available for the backup to continue. You can create customized notifications to change and expand how the NetWorker software behaves when an "AFTD Filesystem Full" notification occurs. Custom notifications can also run custom scripts and other programs to expand the capacity of existing AFTDs.

The chapter "Reporting NetWorker Datazone Activities" provides more information about how to configure notifications.

Creating a custom notification to extend disk space

While the NetWorker default

Filesystem Full — Recover adv_file Space notification works by removing its expired save sets, a custom notification could be configured to expand disk or file system space in other ways.

Procedure

1. In the server’s Administration interface, click Server.

2. Right-click Notifications and select New.

3. For Name, type a unique name for this custom notification, such as First

adv_full notice.

4. For Event, clear all choices except adv_file.

5. For Priority, clear all choices except Waiting.

6. For Action, specify the full path of the custom script that is configured to expand diskspace, for example: /mybin/my_first_custom_script.

7. Click OK.

Creating a custom notification for insufficient disk space

The NetWorker default

Filesystem Full — Waiting for adv_file Space notification works by sending an email notification. A custom notification could be configured to

120 EMC NetWorker 9.1 Administration Guide

Backup Storage do whatever the user indicates. The wait time after the default notification is approximately 10 minutes.

Procedure

1. In the server’s Administration interface, click Server.

2. Right-click Notifications and select New.

3. For Name, type a unique name for this second custom notification, such as

Second adv_full Notice.

4. For Event, clear all choices except adv_file.

5. For Priority, clear all choices except Critical, Emergency, and Alert.

6. For Action, specify the full path of the custom script to be run, for example: / mybin/my_second_custom_script.

7. Click OK.

AFTD load balancing

You can adjust the target and max sessions attributes per device to balance the data load for simultaneous sessions more evenly across available devices. These parameters specify the maximum number of save sessions to be established before the NetWorker server attempts to assign save sessions to another device.

For AFTDs, all volumes, depending on the selection criteria (pool settings), choose the

AFTD with the least amount of data written to it, and join sessions based on the device's target and max sessions. If the number of sessions being written to the first device exceeds the target sessions setting, another AFTD is considered for new backup sessions and is selected from the remaining suitable AFTDs. The AFTD that is selected will be the AFTD with the least amount of NetWorker data written to it. The least amount of data written is calculated in bytes (not by percentage of disk space used) and only bytes that were written by NetWorker are counted.

To ensure that a new session always writes to the AFTD with the least amount of data written to it, you can set each AFTD device's max sessions attribute to 1. However, setting the max sessions attribute to 1 may not be practical. Alternatively, set the target sessions attribute to 1. In this way, load balancing will occur on a best efforts basis.

Space management for AFTD

A configurable setting for determining at what capacity the NetWorker software should stop writing to an AFTD spans from 1 to 100%. Setting the value to 0 or leaving the attribute empty in the AFTD Percentage Capacity attribute is equivalent to a setting of 100%. This means that the entire capacity of the file system can be used for the AFTD volume.

When set, the AFTD Percentage Capacity attribute is used to declare the volume full and to calculate high/low watermarks. When the percentage capacity attribute is modified, mount and re-mount the volume for the new settings to take effect.

The level watermark is calculated based on the percentage of restricted capacity, not on the full capacity of the file system.

In the Console Administration interface, the AFTD Percentage Capacity displays in the

Configuration tab of the Properties window of a device, when Diagnostic Mode is enabled.

To enable Diagnostic Mode, select View

Diagnostic Mode.

Advanced file type devices 121

Backup Storage

NOTICE

If your device uses compression or deduplication, you can still use the AFTD

Percentage Capacity attribute however, the device will be marked as having reached its threshold prematurely. In this case, there will be more unused space on the disk than expected. This is because the threshold limit is based on the amount of data being protected without accounting for the effect of compression or deduplication.

AFTD operation verification

The AFTD can be deployed in varying environments with local disks, and with NFSmounted or CIFS-mapped disks. The configuration of this feature affects its operation. Ensure that the AFTD is fully operational in the production environment before deploying it as part of regularly scheduled operations.

As part of the validation process, test these operations: l l l l

Backup

Recover

Staging

Cloning l l

Maximum file size compatibility between the operating system and a disk device

Use of a volume manager to increase the file system size while the file system is in use l

File system behavior when the disk is full

Some versions of NFS and CIFS drop data when a file system becomes full. Be sure to use versions of NFS, CIFS, and operating systems that fully support full file systems.

On some disk devices, the volume labeling process can take longer than expected.

Labeling time depends on the type of disk device used and does not indicate a limitation of the NetWorker software. The upper limits of save set size depend on either the upper limits supported by the operating system or the file size specified by the disk device’s vendor.

NOTICE

Do not edit device files and directories. This can cause unpredictable behavior and make it impossible to recover data.

Deactivate and erase an AFTD

You can deactivate an AFTD device so it does not interfere with normal backup operations.

Converting a device to read-only

Conversion of a device to read-only prevents the use of the device for backup operations. The device can still be used for read operations, such as restore and clone.

Procedure

1. In the NMC window for the NetWorker server, click the Devices view and select the Devices folder in the navigation tree.

2. In the Devices table, right-click the device to be converted to read-only, and select Unmount.

3. Right-click this unmounted device and select Properties.

122 EMC NetWorker 9.1 Administration Guide

Backup Storage

4. In the Device Properties window, select Read only, and click OK.

5. Right-click the device and select Mount.

Disabling a device

Disabling a device prevents further operation of the device. The device may be reenabled to restore old data, which is retained but not active.

Procedure

1. In the NMC window for your NetWorker server, click the Devices view and select the Devices folder in the navigation tree.

2. In the Devices table, right-click the device to be disabled and select Unmount.

3. Right-click this unmounted device and select Enable/Disable to disable.

4. Inspect the Enabled column of the table to verify that the device is disabled.

Deleting a device

The procedure for deleting a device includes an option for also erasing the volume

(access path) that stores the device’s data. The volume can be erased only if no other device in the system shares the volume.

Procedure

1. In the NetWorker server Device view, click Devices in the navigation tree.

2. In the Devices table, right-click the device to be removed and select Delete.

A confirmation window appears.

3. In the confirmation window: l l

To delete the device from the NetWorker configuration only, without erasing the device’s data, click Yes.

To delete the device and erase the device’s data and volume access path, select the Permanently erase all data and remove media and index

information for any selected AFTDs or Data Domain devices option, and click Yes.

Note

If the volume that you want to erase is shared by another device, then an error message displays the name of the other device. You must delete all other devices that share the volume until the last one remaining before you can erase the volume.

4. If the device is mounted or the device is a member of a pool, then a second confirmation window displays the details of the device and pool. To confirm the device unmount, the removal of the device from the pool, and the deletion of the device, click Yes.

Concurrent AFTD recovery operation limitations

AFTD concurrent recovery currently has the following limitations: l l

Not available to the Windows recover interface (winworkr). Use the recover command. The EMC NetWorker Command Reference Guide or the recover man page provides more information.

Not available to nonfile recoveries, such as NDMP and NetWorker database modules.

Advanced file type devices 123

Backup Storage l

Perform concurrent recoveries from the command line by using the recover command, either by using multiple -S options to identify multiple save sets, or running multiple recover commands concurrently.

When you recover data from an AFTD, NetWorker recovers the save sets concurrently. You can recover multiple save sets to multiple clients simultaneously and you can clone save sets from an AFTD to two different volumes simultaneously.

Changing the AFTD block size

The maximum potential block size for backups to an AFTD device can be adjusted.

Larger block sizes for backups can improve backup speed under certain conditions.

This is especially noticeable on remote AFTD devices that are not local to the storage node, for example, AFTDs that are connected with CIFS or NFS.

Changes to the maximum potential block size value for an AFTD device take effect only after the AFTD device is labelled. The minimum allowable block size is 128 kilobytes and the maximum block size is 256 kilobytes.

If you have an AFTD device that is performing backups slowly, try marking the device as read-only and create a new AFTD device with a block size between 128-256 kilobytes.

NOTICE

Changing the block size and re-labeling an existing AFTD has the potential to destroy data if the data is not staged to another location.

Procedure

1. In the server’s Administration interface, click Devices.

2. Select View

Diagnostic Mode.

3. Select Devices in the navigation tree. The Devices detail table appears.

4. Double-click the device in the devices table and select the Advanced tab.

5. In the Device block size attribute, select a value from 128 to 256.

6. Click OK.

7. Relabel the AFTD device for the new setting take effect.

DD Boost and Cloud Tier devices

DD Boost and Cloud Tier devices are covered separately in the EMC NetWorker Data

Domain Boost Integration Guide

Creating a DD Boost device

Procedure

1. In NMC, click Devices.

2. In the left panel, right-click Devices and select New Device Wizard.

3. On the Select the Device page, select Data Domain and click Next.

4. On the Data Domain Preconfiguration Checklist page, click Next.

5. On the Specify the Data Domain Configuration Options page: a. Under Data Domain System Name: l

Select Create a New Data Domain System.

124 EMC NetWorker 9.1 Administration Guide

Backup Storage l

In the text box, type the IP address of the Data Domain system.

b. In the Data Domain DDBoost Username field, type the username of the

Data Domain user.

c. In the Data Domain DDBoost Password field, type the password of the

Data Domain user.

d. Specify the required values in the other fields.

e. Click Next.

6. On the Select the Folder to Use as Devices page: a. Click New Folder to create a folder for the device.

b. Select the newly created folder.

c. Specify the required values in the other fields.

d. Click Next.

7. On the Configure Pool Information page: a. Under Pool Type, select one of the following pool types: l l

Backup

Backup Clone b. Under Pool, perform one of the following tasks to select the pool: l l

Select Create and use a new pool, and type the pool number in the text box.

Select Use an existing pool, and select the pool from the drop-down list box.

c. Specify the required values in the other fields.

d. Click Next.

8. On the Select Storage Nodes and Fibre Channel Options page: a. Select the storage node.

b. Specify the required values in the other fields.

c. Click Next.

9. On the Select SNMP Monitoring Options page, specify the required field values, and click Next.

10. On the Review the Device Configuration Settings page, review the configuration settings, and click Configure.

11. On the Device Configuration Results page, click Finish.

Atmos Cloud devices

This section describes how to configure the NetWorker Cloud Backup Option (NCBO) to perform backup, staging, cloning, and recovery operations to cloud configurations.

Backups to cloud occur over a TCP/IP network and can be compressed and encrypted. NetWorker supports EMC Atmos-based cloud storage.

Atmos Cloud devices 125

Backup Storage

Cloud backup requirements and considerations

The following conditions must be met before you can backup to the cloud.

l l l

The NetWorker Cloud Backup Option must be licensed and enabled. The EMC

NetWorker Licensing Guide provides information about license enablers.

The NetWorker Cloud Backup Option is supported on Windows and Linux storage nodes only.

An Atmos cloud account is set up and you have a username and password to access the cloud account. The EMC Atmos Installation Guide and the EMC Atmos

System Management GUI Guide provides information about setting up and managing an Atmos account.

If the Atmos server and the NetWorker server are separated by a firewall, TCP ports 80 and 443 must be open to allow outgoing communication from the

NetWorker server to the Atmos server. If a proxy server is configured in the environment, a firewall exception may also need to be created to ensure unrestricted access. If these ports are not open, device operations will fail with the following error: Atmos label operation failed: Failed to write cloud label: Couldn't connect to server.

For NDMP, only a Data Server Agent (DSA) is supported for cloud backups.

Cloud best practices

Consider the topics and recommendations in this section before implementing cloud backups.

Backups to a cloud storage device

Backups are sent to a cloud storage device by using media pools, in the same way that client backups are directed to a device or set of devices.

Consider the following recommendations: l l

Set up one or more media pools for CloudBoost devices and provide the pool with a unique label template.

Do not mix CloudBoost backup devices with other types of backup devices in a media pool.

Concurrent backup and recovery operations

You can mount a single CloudBoost appliance as multiple cloud storage devices to support concurrent backup and recovery operations.

For example, to optimize performance you can mount the CloudBoost appliance on three cloud storage devices: one for backup (device CL1), one for recovery (device

CL2), and one for clone operations (device CL3). There is no limit to the number of cloud storage devices that can be mounted on a single cloud volume. Consider such an approach to optimize backup and recovery performance.

Network dependencies

Cloud backups are highly dependent on the network connection that is used to access the cloud service. Any disruption in connectivity or a slowdown in network access speed may adversely affect cloud backups or recoveries.

Proper name resolution and internet access is required for the CloudBoost appliance.

The CloudBoost appliance documentation provides more information.

126 EMC NetWorker 9.1 Administration Guide

Backup Storage

HTTPS communications with the Cloud server

NetWorker enables you to use https to securely communicate with the Cloud server.

Configure https communications when you configure the cloud device in NetWorker.

You will need to retrieve the SSL certificate for the Cloud server before you create a new device or modify an existing cloud device. Use the nsrssltrust command on the NetWorker server to retrieve the SSL certificate of the Cloud server and store the certificate to a local file.

To retrieve the certificate file, type the following command: nsrssltrust -c path\filename -u https_server where: path\filename is the location to store the certificate file on the NetWorker server.

https_server is the hostname of the https server.

For example: nsrssltrust -c ./atmos1.cert -u https://accesspoint.atmosonline.com:

443

Create and label a cloud storage device

It is required to create and label cloud storage devices for use with the NetWorker software.

Creating the cloud storage device

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Right-click Devices in the navigation tree, and select New. The Create Device window appears, with the General tab selected, and a default device path in the

Name field of the Identity area of the window.

3. In the Name field, replace the default name with a name that uniquely identifies the cloud storage device. If the device is configured on a remote storage node, indicate that the storage node is remote, by including rd=hostname: in the name. For example, if the remote storage node is neptune, then the device path might be rd=neptune:cloud1.

Note

A cloud storage device name does not specify a path to the device. You can use any combination of alphanumeric characters for the device name.

4. In the Comment field and the Description field, add an optional comment and description, respectively.

5. In the Media Type field, select Atmos COS as the device type if you are using

Atmos as the cloud server.

6. In the Remote User field, type the username that is used to access the cloud server. For an Atmos COS device, this is the token-id.

7. In the Password field, type the password that is used to access the cloud server. For an Atmos device, this is the shared secret.

Atmos Cloud devices 127

Backup Storage

128

8. Select the Cloud tab to specify additional information specific to the cloud backup device.

9. In the Server field, type the IP address or fully qualified domain name of the cloud server.

10. Use the Parameter options to adjust network communication attributes: a. In the Network Write Size field, specify the amount of backup data, in kilobytes, to cache in memory before sending to the cloud. Larger write sizes typically result in better performance but results vary depending on the underlying network characteristics. Also be aware that larger sizes consume more memory on the storage node for the duration of the backup or recover operation.

b. In the Number of Retries field, specify the number of times that NetWorker will attempt to send backup or receive recover data in the event of a network failure.

c. In the Send/Receive Timeout field, specify the number of seconds that

NetWorker will wait for confirmation that network send and receive transmissions to the cloud server have occurred successfully. If the timeout period expires, the data transmission is considered to have failed.

l l

Set the value of this field in conjunction with the Network Write Size field. Larger Network Write Size values require larger Send/Receive

Timeout values to avoid failures. Optimal values for the Send/Receive

Timeout field vary depending on the network speed and bandwidth.

The save group’s Inactivity Timeout value can potentially interact with the Send/Receive Timeout value in unintended ways. To avoid this possibility, ensure that the save group’s Inactivity Timeout value

(default is 30 minutes) is greater than the Send/Receive Timeout value

(default is 30 seconds).

d. In the Network Failure Retry Interval field, specify the number of minutes that a backup or recover session must wait before a failed network connection results in an aborted backup or recover session.

11. In the Compression field, select a compression level for data that is sent to the cloud. Faster compression speeds result in less data compression but also require less CPU resources. The fastest compression speed, Compression

Speed Fast, performs the least amount of data compression and is selected by default.

To choose an optimal compression value, balance the potentially longer backup window of using a slower compression speed against the potential efficiency and cost savings of sending less backup data to the cloud.

NOTICE

If the NetWorker Cloud Back Option determines that backup data cannot be compressed effectively, compression may not occur regardless of the setting in this field.

12. In the Encryption field, specify whether to enable or disable encryption of data sent to the cloud. Encryption is standard NetWorker AES 256 bit encryption and is selected by default. If desired a NetWorker datazone pass phrase can be defined that would be used to recover encrypted data.

If this option is selected, encryption will occur regardless of any client-side encryption directives. The EMC NetWorker Security Configuration Guide provides

EMC NetWorker 9.1 Administration Guide

Backup Storage more information about encryption including how to specify a new datazone pass phrase.

NOTICE

If encryption is already enabled for the NetWorker client and encryption is enabled in this field, backups will be slower because encryption functions will occur twice.

13. Use the Cloud network interface field if the Storage node has multiple network interfaces. If it does, specify the IP address of the network interface that will send backup data to the cloud.

To display the Cloud network interface field, select View

Diagnostic Mode

from the menu bar.

14.

Select Throttling and then click the Bandwidth icon to display a dialog box where you can adjust the maximum internet bandwidth that a cloud backup or recovery operation can consume at any given time of the day or week. This option enables you to prevent network congestion by limiting cloud backup and recovery activity during peak internet usage.

a. Select New to add a bandwidth throttling policy.

b. From the Day field, select the day of week to which the policy applies.

c. Click the up and down arrows to select a time of the day to which the policy starts and ends. Alternatively, type the times directly into the Start time and End time fields.

d. Click the up and down arrows to select the maximum possible network bandwidth. in megabits per second, that a backup or recovery operation can consume when the policy is in effect. Alternatively, type the values directly in the fields.

You can create as many policies per day as required. You can also modify or delete existing throttling policies as necessary.

15. Click OK when the configuration is complete.

Labeling and mounting the cloud storage device

If there are multiple cloud volumes, you will be able to select the volume to associate with the cloud storage device.

Procedure

1. Select the cloud storage device, right-click and select Label. The Create new

cloud volume dialog box appears.

2. In the Pools field, select the media pool to be used for cloud storage devices.

Note

It is recommended that the media pool you select be used for cloud backup devices only.

A label for the cloud storage device is generated and displayed in the Volume

Label field. The label name is based on the label template that was specified for the cloud media pool.

Atmos Cloud devices 129

Backup Storage

3. Select Mount after labeling and click OK.

Report information on cloud backup

Use cloud backup information to monitor backup costs and help optimize your cloud backups.

Cloud backup information can be obtained from the following sources: l l

Cloud backup and recover reports in NMC.

The mminfo command

Use the mminfo -avot command to get information on how much data is consumed in a cloud backup. The EMC NetWorker Command Reference Guide and the UNIX man pages provide more information about how to use the mminfo command.

Cloud storage device staging

Staging with a cloud storage device works the same way as staging to a tape device.

You cannot however, use a cloud storage device as the source for a staging operation.

Staging save sets

on page 444 provides more information.

Cloud storage device cloning

Cloning with a cloud storage device works the same way as cloning with any other advanced file type device.

Libraries and silos

NetWorker supports SCSI libraries, NDMP libraries, and ACSLS silos. In a fibre channel environment you can configure library and device sharing between storage node hosts.

Overview of tape device storage

This chapter contains information on the creation, configuration, and management of tape devices. Tape devices may be configured as stand-alone devices or configured as part of a traditional tape library or virtual tape library (VTL) storage system.

The libraries and devices available to a NetWorker server are listed in the Devices view of the NetWorker Administrator window. The details and settings of a particular device can be viewed by right-clicking the device and selecting Properties. The full range of property attributes can be viewed by selecting View > Diagnostic Mode. A description of the various attributes is provided by the Field Help button.

As with other Console functions, you can view and work with only those NetWorker servers for which you have access permission.

NetWorker software supports many different types of tape libraries, also called autochangers or jukeboxes. The general categories of libraries are SCSI, NDMP, and silo.

Support for LTO-4 hardware-based encryption

The use of LTO-4 hardware-based encryption is supported by NetWorker when controlled by management utilities that are provided with the LTO-4 hardware, or by third-party key management software. EMC does not test or certify these key management utilities; however, the NetWorker application can read from and write to

LTO-4 devices that use hardware-based encryption. The use of this encryption is

130 EMC NetWorker 9.1 Administration Guide

Backup Storage transparent to NetWorker. Neither the encryption nor the key management process is managed by the NetWorker application. This includes the ability to turn encryption on or off within NetWorker, and the management of encryption keys.

Linux device considerations

Review this section for information about using devices on Linux hosts.

Configure Linux operating system to detect SCSI devices

Proper configuration of the SCSI subsystem is required to get full use of SCSI devices and allow the operating system to detect SCSI devices that are attached to the computer. If the device is configured with multiple LUNs, set the kernel parameter

Probe all LUNs of each SCSI Device to Yes.

The Linux Documentation Project website provides more information on configuring the Linux SCSI subsystem. For information on the SCSI device, contact the manufacturer.

The inquire command and the Scan for Devices operation do not detect more than 128 tape devices

By default, the Linux st kernel module only configures up to 128 SCSI tape devices

(/dev/nst).

When the number of SCSI tape devices exceeds the kernel value ST_MAX_TAPES, the following error may appear in the /var/log/messages operating system log file: st:Too many tape devices (max. 128)

The inquire command or the Scan for Devices option in NMC only displays the maximum number of st devices (/dev/nst) defined by the ST_MAX_TAPES value.

To resolve this issue, edit and recompile the st module of the Linux kernel to increase the maximum number of allowable st devices that are created by the OS to exceed the default value. The Linux documentation provides details on how to reconfigure, rebuild, and install the kernel.

Configuration requirements for the inquire command

Depending on the specific OS requirements and the configuration of the NetWorker server or storage node, you may need to create device files so that the inquire command can detect all devices.

For example, on a NetWorker server with Red Hat Linux, if devices sg0 through sg15 exist, create device file sg16 by using the mknod program as follows:

mknod /dev/sg16 c 21 17

The operating system vendor documentation provides more information on creating devices.

Linux device considerations 131

Backup Storage

Solaris device considerations

Review this section for information about using devices on Solaris hosts.

Support for tape devices not supported by Solaris

If Sun Microsystems does not directly support a device for use with the operating system on the storage node, obtain a st.conf file from the device manufacturer.

The inquire command and Solaris 10

On Solaris 10, the inquire command does not show library information after you configure the library for NetWorker.

HP-UX device considerations

Review this section for information about using devices on HP-UX hosts.

Autochanger installation on an HP-UX system

The following sections explain how to install and configure Hewlett-Packard drivers.

Selecting SCSI addresses for the autochanger

Determine which SCSI address is assigned to each SCSI bus, and select the SCSI addresses to be allocated to the autochanger drives and controller.

To select unused SCSI addresses for an autochanger, log in as root on the NetWorker server or storage node, and type the ioscan -f command.

Use a SCSI address within the range of 0 to 6. The primary hard disk is usually on

SCSI address 6.

132

NOTICE

For some devices, such as the HP Model 48AL autochanger, select one SCSI address for the entire autochanger. The 48AL uses a different SCSI logical unit number (LUN) for the device (LUN 0) and robotics (LUN 1). The SCSI LUN appears as the last digit of the H/W Path field in the ioscan output.

Installing the SCSI pass-through driver

The following procedure describes how to use SAM terminal mode to install a GSC,

HSC, or PCI pass-through driver.

Procedure

1. Select Kernel Config and press Enter.

2. Select Drivers and press Enter.

3. Select the SCSI_ctl driver by selecting SCTL from the list.

If the current state is in, go to step 9. Otherwise, select any unreserved name for the device. For example, do not select a name such as /dev/null.

4. From the Actions menu, select Add Drivers to Kernel, and press Enter.

5. From the Actions menu, select Create a New Kernel, and press Enter.

A confirmation message appears.

6. Specify Yes, and press Enter.

The Creating Kernel message appears, followed by the Move Kernel message.

EMC NetWorker 9.1 Administration Guide

Class spt

Backup Storage

7. Select OK, and press Enter.

The system reboots.

8. Verify that the spt was successfully installed by typing the following command:

ioscan -kfn

9. Verify that the driver has claimed the autochanger.

If the autochanger has been claimed, CLAIMED should appear under the S/W

State header. If not, verify that the installation completed correctly.

10. If the device entry was defined by the operating system, use the OS-defined entry and continue to verify the installation.

Determining the major number

To determine the value for majornum, type lsdev -d sctl.

The output should resemble the following example output, although the assigned number may differ from the values in this example:

Table 28 Determining the major number value

Character

HP-PB

HSC or PCI

Block

75 -1

203 -1

Driver spt sctl

Class spt ctl

The value for majornum is the number in the Character column.

Determining the minor number

To determine the value for minornum, use the ioscan command.

The relevant lines in the ioscan output are those: l

For the controller itself, which contains HP C6280-7000 in the Description column.

For the adapter to which the controller is connected, which is the second line above the line for the controller and contains ext_bus in the Class column.

If the schgr driver is configured on the system, it appears associated with the library.

The ioscan output line resembles:

Table 29 ioscan output when driver is configured

H/W Path

10/4/4.6.0

Driver schgr

S/W State

CLAIMED

H/W Type

DEVICE

Description

C6280-7000

If the schgr driver is not configured on the system, no driver appears to be associated with the library. The ioscan output line resembles:

HP-UX device considerations 133

Backup Storage

Class unknown

-1

Table 30 ioscan output when driver is not configured

H/W Path

10/4/4.6.0

Driver schgr

S/W State

UNCLAIMED

H/W Type

DEVICE

Description

C6280-7000

Testing the device driver and device file installation

After the device driver is installed and the device file is created, run the inquire command to list available SCSI devices.

NOTICE

Use the inquire command with caution. Running inquire sends the SCSI inquiry command to all devices detected on the SCSI bus. Using the inquire command during normal operations may cause unforeseen errors and possible data loss may result.

An example of the output from this command (with the -s option) is as follows: [email protected]:HP C1194F 0.14 Autochanger (Jukebox), /dev/rac/ c0t1d0 [email protected]:Quantum DLT4000 CC37 Tape, /dev/rmt/c0t2d0BESTnb [email protected]:Quantum DLT4000 CC37 Tape, /dev/rmt/c0t3d0BESTnb [email protected]:Quantum DLT4000 CC37 Tape, /dev/rmt/c0t4d0BESTnb [email protected]:Quantum DLT4000 CC37 Tape, /dev/rmt/c0t5d0BESTnb

As of HP-UX 11iv3, two different addressing modes are supported: LEGACY and

AGILE. The inquire program lists devices using the B.T.L. notation for the LEGACY addressing mode, for example: [email protected].

For the AGILE addressing mode, it lists devices using the DSF notation, for example:

/dev/rtape/tape106_BESTnb

Inquire command does not detect tape drive

When a tape drive is attached to the HP-UX 11i V2 64-bit host and the inquire command is run, the tape drive is not detected, even if the device is configured, labeled, and mounted and a save was successful.

To work around this issue, identify the drive path in the /dev/rmt folder, and configure the device with this path.

Whenever a new device is attached to the system, ensure that the cached file /tmp/ lgto_scsi_devlist is updated. Remove this temp file and then run the inquire command, which rebuilds the file.

Errors from unsupported media in HP tape drives

Certain HP tape drives can only read 4-mm tapes of a specific length. Some, for example, read only 60-meter tapes. To determine the type of tape that is supported, refer to the drive’s hardware manual.

If unsupported media is used, the following error message may appear when you use the nsrmm or nsrjb command to label the tape:

134 EMC NetWorker 9.1 Administration Guide

Backup Storage nsrmm: error, label write, No more processes (5)

The following error message may appear when you use the scanner -i command to label the tape when unsupported media is used: scanner: error, tape label read, No more processes (11) scanning for valid records … read: 0 bytes read: 0 bytes read: 0 bytes

Unloading tape drives on an HP-UX server or storage node

When the nsrjb -u -S command is used to unload a tape drive in an autochanger that is attached to an HP-UX server or storage node, the unload operation ejects all tape volumes inside the autochanger devices, and into their respective slots.

To unload a single drive to its corresponding slot, use the nsrjb -u -f device_name command instead.

SCSI pass-through driver required for HP-UX autochangers

Review the required procedures in the EMC NetWorker Installation Guide before you run the jbconfig program to configure an autochanger with a NetWorker server on

HP-UX.

Follow the procedures to rebuild the kernel even if the SCSI pass-through driver is installed. Then run the jbconfig program to configure the autochanger.

AIX device considerations

Review this section for information about using devices on AIX hosts.

STK-9840 drives attached to AIX

If you attach an STK-9840 drive to an AIX server, use SMIT to modify the IBM tape drive definition field to set the value of Use Extended File Mark to Yes.

LUS driver operation on AIX

When a library comes online, NetWorker obtains an exclusive lock on the library due to the operation of the LUS driver on AIX. This lock is maintained if the library is enabled.

As a result, you cannot use diagnostic tools such as inquire and the sji utilities to access the library during this time. To access the library using these tools, you must first take the library offline.

SCSI and VTL libraries

SCSI libraries have automated robotic mechanisms to move tape media from a fixed number of library slots to devices for read or write operations. The number of slots can typically vary between 2 to 10,000 and the number of devices can be between 1 to

100 or more.

Traditionally, libraries are physical units with mechanical robotics, however the same functionality can also be provided by virtual tape libraries (VTLs) that emulate this functionality. VTLs can also be configured and used as Autochangers.

The robotic controller and associated tape devices are always all controlled through a

SCSI interface which is available on one or more storage hosts.

AIX device considerations 135

Backup Storage

Selecting a volume for the NetWorker server

When a backup takes place, the NetWorker server searches for a volume from the appropriate pool to accept the data for backup.

The available volumes are as follows: l l l

Mounted on stand-alone devices.

Available for labeling and accessible to the NetWorker server through Auto Media

Management or a library.

Labeled for the appropriate pool and already mounted in a device, or are available for mounting, if a library is being used.

If two or more volumes from the appropriate pool are available, the server uses this hierarchy to select a volume.

l l l l l

A volume in a jukebox device has priority over volume in a disk or tape device.

A volume in a local disk device has priority over a volume in a local tape device.

If two local disk are available, then the device less data sessions will have priority.

if two local tapes devices have available volumes, then NetWorker will use the volume with the earliest label date.

If two jukebox are available, then NetWorker will select the volume with the earliest label date..

Data recovery and volume selection

The NetWorker server determines which volumes are required for recovery. If the appropriate volume is currently mounted, the recovery begins. If the volume is not mounted and a library is used, the server attempts to locate and mount the volume in an eligible device for appropriate media pool. Preference is given to mount the volume in a read-only device, if one is available.

If a stand-alone device is used, or if the server cannot locate and mount the volume, the server sends a mount request notification.

If more than one volume is needed to recover the data, the NetWorker server displays all the volumes, in the order needed. During the recovery process, the server requests the volumes, one at a time.

NOTICE

NetWorker will automatically unload volumes that have been placed in a jukebox device but have never been mounted (for example, nsrjb -l -n <volume>). Any command, such as the scanner command, that operates on volumes that have never been mounted will be affected by this behavior. To prevent NetWorker from unloading the volume, the device should be set to service mode while the command is being run.

Automatic volume relabeling

NetWorker has the ability to automatically relabel recyclable volumes when needed or when scheduled.

When you enable Auto Media Management, the NetWorker server will automatically relabel a volume when the mode is recyclable. A volume is automatically set to recyclable when all save sets on the volume, including partial save sets that span other volumes, are marked as recyclable.

Auto Media Management

on page 146 provides more information on Auto Media Management.

136 EMC NetWorker 9.1 Administration Guide

Backup Storage

Note

You can manually change the mode of a volume to recyclable.

Changing the volume mode on page 472 provides information about changing the mode of a volume.

You can configure a media pool to automatically relabel recyclable volume at a user

defined time and interval. Automatically relabeling volumes in a media pool on page 87

provides more information about configuring the automatic relabel process for recyclable volumes in a media pool.

Virtual tape library (VTL) configuration

During library configuration, the NetWorker software automatically attempts to detect if a library is a VTL, and updates the read-only Virtual Jukebox attribute to Yes, or if not, to No. VTLs that are mistakenly identified as autochangers can indicate what type of license should be used, either autochanger or VTL.

VTL licensing

The EMC NetWorker Licensing Guide provides information about NetWorker licensing support for a Virtual Tape Library.

Multiplex backups to Data Domain VTL devices

You can configure multiplexed backups to Data Domain VTL devices on remote, nondedicated NetWorker storage nodes. Multiplexing is the use of multiple parallel save streams or concurrent sessions to each device. Each additional save stream (max sessions value) to a VTL device reduces the number of devices needed by somewhat less than one because deduplication efficiency decreases slightly.

The following prerequisites, restrictions, and considerations apply: l l l l l

NetWorker dedicated storage nodes (DSNs) and NetWorker backup to local VTLs cannot use this configuration.

If you are currently using DD OS 5.0.x, upgrade to DD OS 5.0.2 or later.

If you are currently using DD OS 4.x, upgrade to DD OS 4.9.3.1.

Multiplexing decreases deduplication efficiency on the VTLs by 4% to 8% per additional save stream. For example, given a sufficiently large device block size, 4 parallel streams (max sessions=4) results in deduplication ratios that are 12%-24% below the non-multiplexed rate (max sessions=1).

Deduplication ratios may be initially low when you increase max sessions due to extra processing, following which efficiency improves.

l l

Heavily used Data Domain systems, with 75% or more disk space already used, can suffer impaired performance when used with multiplexing.

As a best practice, do not use client-side or server-side encryption during backup to the Data Domain system.

Multiplex to Data Domain VTL prerequisites and considerations

Ensure the following prerequisites and practices.

l l l

If currently using DD OS 5.0.x, upgrade to DD OS 5.0.2 or later.

If currently using DD OS 4.x, upgrade to DD OS 4.9.3.1.

The recommended settings for VTL are: max sessions=4; target sessions=4; and device block size=512 KB.

SCSI and VTL libraries 137

Backup Storage

138 l l l l l

Best max sessions and device block size values depend on the environment. For example, max sessions=2 might provide better stability and deduplication while still meeting the backup window.

Deduplication efficiency on the VTLs is reduced by 4% to 8% per additional save stream. For example, given a sufficiently large device block size, 4 parallel streams

(max sessions=4) results in deduplication ratios that are 12%-24% below the nonmultiplexed rate (max sessions=1).

Typically, deduplication ratios are initially low when you increase max sessions and device block size due to re-priming and re-analysis overhead, following which efficiency improves.

Heavily used Data Domain systems, with 75% or more disk space that is already used, can suffer impaired performance when used with multiple sessions.

As a best practice, do not use client-side or server-side encryption during backup to the Data Domain system.

Configuring multiplex backup to Data Domain VTL devices

Configure Data Domain VTL devices for multiple session backups as follows.

Procedure

1. Shut down backup service on the NetWorker VTL storage node, or shut down the NetWorker server if that is possible, and verify that there is no backup activity on the storage node.

2. Use NMC or the nsradmin command to set the sessions values for each VTL device. The recommended values are: max sessions=4 (32 maximum); target sessions=4; and device block size=512KB.

Optimal max sessions and device block size values depend on your environment.

For example, max sessions=2 might provide better stability and deduplication while still meeting your backup window.

Note

If you shut down the NetWorker server in step 1, you can run the nsradmin command with the -d resdir option. This option uses the NetWorker resource database, resdir , without opening a network connection.

For example, on UNIX/Linux or Microsoft Windows systems, run the following command:

nsradmin -i input_file.txt

where input_file.txt is a text file that contains the following lines that you can customize to your own environment:

option regexp: on

. type: nsr device; media type: LTO Ultrium-3; media family: tape; name: /dev/rmt* update max sessions: 4; target sessions: 4; device block size:

512KB

3. Create a no intra-block multiplexing (nibmp) tag file in the NetWorker debug folder on the NetWorker storage node.

For example, you can use the standard NetWorker installation paths for the tag file. You can limit the tag file path to a specific pool by adding the _poolname variable as a suffix to the tag file. The _poolname can include spaces, for

EMC NetWorker 9.1 Administration Guide

Backup Storage example, _My Pool. On Microsoft Windows systems, ensure the specified pathname is enclosed in quotes.

Unix/Linux system examples.

touch /nsr/debug/nibmp touch /nsr/debug/nibmp_My Pool

Microsoft Windows system examples.

echo > "NetWorker_install_path\nsr\debug\nibmp" echo > "NetWorker_install_path\nsr\debug\nibmp_My Pool"

4. Restart the NetWorker services to enable the multiplexing functionality.

The technical note named NetWorker Improved Deduplication with Multiplexing to

Data Domain VTLs Technical Note , available on the EMC Online Support website, provides more details.

Non-rewinding tape device usage (UNIX/Linux only)

Tape drives used as storage devices must be accessed by non-rewinding device files.

The NetWorker server assumes that a tape is in the same position in which it was the last time it was accessed. If the operating system’s device driver rewinds the tape, then the position is lost, and previously written data will be overwritten by the next backup.

The NetWorker configuration software automatically chooses the correct device pathname for tape devices. If the user specifies the pathname, then it must be nonrewinding, and it must follow the Berkeley Software Distribution (BSD) semantic rules.

For example, /dev/rmt/0mbn, where: l

The b satisfies the BSD semantics requirement on Solaris and HP-UX.

The n specifies non-rewinding behavior on Solaris, HP-UX, Linux, and HP-Tru64.

On AIX, the number following the decimal selects the BSD and non-rewinding behavior and must be either 1 or 5 for NetWorker software (for example /dev/rmt2.1).

Note

Never change a device pathname from non-rewinding (/dev/rmt/0cbn) to rewinding (/dev/rmt/0cb). When the pathname is changed to rewinding, the data could only be saved, but never recovered. All but the last save are overwritten by later saves.

Pools with libraries

If the backup strategy includes both full and nonfull backups, estimate the number of volumes needed for the full backups and assign them to the Full pool. This ensures that the full backups are located in a consecutive range of slots in the library. This allows all of the volumes to be removed at the same time.

Persistent binding and naming

Some operating systems provide the persistent binding option to permanently bind logical and physical addressing so that the associations are retained. This guarantees

SCSI and VTL libraries 139

Backup Storage that the operating system always uses and creates the same symbolic path for a device is known as persistent naming.

Proper configuration of the operating system to use persistent binding and persistent naming resolves issues related to device ordering by forcing the operating system to always assign the same device filename regardless of external events.

Persistent binding

Persistent binding guarantees that the operating system always uses the same SCSI target ID for SAN devices, regardless of reboots or other events, by statically mapping a target's WWN address to a desired SCSI address. On some operating systems, this is done by default, while on others it has to be set manually. The operating system documentation provides further information.

In most cases, persistent binding should also be set on the Host Bus Adapter (HBA) by using the configuration utility that comes with the Fibre Channel HBA. The HBA device driver documentation provides details.

Persistent binding is required for consistent library operations within NetWorker, because the NetWorker server communicates with the library controller over a SCSI address that is chosen during initial library configuration. If the SCSI address changes, the library will become unavailable. In this case, disable the library and change the

“control port” address to reflect the new SCSI address of the library controller.

If devices have already been configured in NetWorker prior to enabling persistent binding on the host, delete existing devices from the library resource and perform a re-scan of devices followed by a reconfiguration of the tape library.

Persistent naming

Persistent naming is used to ensure that the operating system or device driver of a server always creates and uses the same symbolic path for a device (referred to as device file).

After you create persistently named device files and they are present on the host, enable the Use persistent names option when scanning for tape devices from the

NetWorker Management Console.

If devices have already been configured in NetWorker prior to enabling persistent naming on the host, delete existing devices from the library resource and perform a re-scan of devices followed by a reconfiguration of the tape library.

Whether to add or recycle volumes

The NetWorker server saves files on volumes marked appen (appendable). If the volumes are marked full, they cannot receive backups. There are situations best suited to either adding a new volume, or recycling an existing volume.

If volumes are marked full, you can: l

Remove the full volumes and replace them with new media if the volumes are being kept for long-term storage.

Change the volume mode to recyc (recyclable) if the data on the full volumes is not needed. The NetWorker server overwrites the data with new backups, but

maintains the existing labels. Changing the volume mode on page 472 provides

information about changing the volume mode.

When all of the save sets on the volume have passed the time period specified by the retention policy, the mode of the volume automatically changes to recyclable.

There are advantages both to recycling media and adding more media to a pool. With recycling, the same volumes are used repeatedly, and there is no need to add new

140 EMC NetWorker 9.1 Administration Guide

Configure libraries

Backup Storage volumes to the pool. The volumes can, however, wear out over time and exhibit a higher failure rate.

On the other hand, if backups are to be stored for some time, then it might be necessary to add more media to the pool instead of recycling. For example, a library might need new volumes every three months if the company policy is to maintain the backups for a year. In this case, new media must be added to the pool until the volumes that contain expired or old backups can be recycled.

A library resource must be created on a storage node for each library, including silos, that you want to use with NetWorker. Because the NetWorker server is also a storage node, this procedure applies to a NetWorker server and all storage nodes. You can configure a library either automatically with the Configure All Libraries wizard or manually with the user interface.

Before you create devices, you must create the storage node that will manage the

devices. Storage nodes on page 93 provides details. When you create the new

devices, you can use NetWorker to perform a device scan, which searches for new devices across multiple storage nodes.

NetWorker can only automatically create tape devices that have serial numbers. Use the inquire or sn commands to determine if a device returns a serial number. UNIX man page and the EMC NetWorker Command Reference Guide provides detailed information about how to use the inquire and sn commands.

NetWorker can automatically configure the following library types: l l l

SCSI

NDMP

ACSLS Silo

Use the jbconfig command to configure a library that contains tape devices or a robotic arm that does not have serial numbers. Use the jbconfig command to configure IBM tape libraries that are controlled through the use of the IBMs tape driver. This is because the device autodetection code uses the internal lus driver to control libraries.

Note

Before you create devices on a storage node, update the devices to the most recent firmware and driver versions.

Autodetection of libraries and tape devices

Autodetection is a scanning process that applies only to physical tape libraries and virtual tape libraries (VTLs). The NetWorker software automatically discovers libraries and devices that are being used for backups and recoveries.

The maximum number of configured devices for any NetWorker server and storage node combination is 750. The maximum number, including non-configured devices, can vary depending on the specific server that is being administered.

The following options are available from many of the menus throughout the Devices task: l l

Configure all Libraries

Scan for Devices

SCSI and VTL libraries 141

Backup Storage

If you start these options from the server folder instead of from the storage node folder, then all storage nodes on the NetWorker server are automatically selected for configuration in the wizard, or for scanning, respectively.

As with other Console functions, you can view and work with only those NetWorker servers for which you have access permission.

NOTICE

Autodetection should not be used for devices on a Storage Area Network (SAN) while any of the devices are in use, because this may cause the device in use to become unresponsive. To avoid this situation, do not configure a device in multiple NetWorker datazones.

Adding a library resource

Procedure

1. In the server’s Administration interface, click Devices.

2. Open the Storage Nodes folder in the navigation tree.

3. Right-click the storage node to which the device is to be configured, and select

Configure All Libraries (which is available from many of the menus throughout the Devices task). This opens a wizard that can configure all detected libraries, except those explicitly excluded in the library exclusion list during configuration.

NOTICE

If Configure All Libraries is started from the server folder instead of from the

Storage Node folder, then all storage nodes on the NetWorker server are automatically selected for configuration in the wizard.

The Configure All Libraries wizard appears. This lets you step through library configuration, including this input (some of which is filled in by default): l l l l

Library type (select SCSI/NDMP).

An NDMP remote username and a password are required for an NDMP device that acts as a storage node.

Adjust the Enable New Device option, if necessary.

Current server sharing policy. Use maximal sharing with Dynamic Drive

Sharing (DDS). By default, the sharing policy is displayed as “server default,” which is maximal sharing.

l l

Storage nodes to which libraries can be configured (select a storage node to see its details). If the appropriate storage node is not listed, click Create a

New Storage Node.

When creating a new storage node, replace the default value in the Name field with the fully-qualified domain name or short name of the new storage node.

Update storage node properties, if required.

4. After specifying the required information, click Start Configuration. The configuration window displays a message that the Configure All Libraries process has started. The status of the configuration activity can be viewed by the Monitoring

Log screen.

142 EMC NetWorker 9.1 Administration Guide

Backup Storage

5. When the configuration is complete, click Finish to close the configuration wizard. If problems occur during configuration, you can click the Back button on the configuration window to adjust the settings.

Scanning for libraries and devices

Devices already known to the NetWorker server can be seen in the enterprise hierarchy in the navigation tree. Use the Scan for Devices option described here to find devices that are not yet known to the NetWorker server. Be aware that: l l l l

A storage node must be added to the hierarchy before its devices can be scanned.

The Scan for Devices option does not detect file type or advanced file type devices.

By default, the Linux kernel configures a maximum of 128 st devices by default.

Refer to

The inquire command and the Scan for Devices operation do not detect more than 128 tape devices

on page 131 if the Scan for Devices option does not detect more than 128 tape devices on Linux operating systems.

A specific network interface can be used between the NetWorker server and the

storage node when scanning for devices. Identifying a specific network interface for device scan operations

on page 145 provides more information.

Procedure

1. In the Console window, click Enterprise.

2. In the navigation tree, select a NetWorker server.

3. In the Name column of the Host detail table, double-click NetWorker. The

NetWorker Administration window for the selected server opens. Note that while multiple NetWorker Administration windows can be open simultaneously, each one displays information about only one host or server.

4. In the Administration window, click Devices.

5. In the navigation tree: a. Right-click the server name, and select Scan for Devices.

b. Click the storage node to be scanned.

c. If the appropriate storage node is not listed, click Create a New Storage

Node.

d. When creating a new storage node, replace the default value in the Name field with the fully-qualified domain name or short name of the new storage node.

e. Fill in any required information, such as whether to scan for SCSI or NDMP devices and whether to search all LUNs.

f. Click Start Scan. To monitor the scan activity, click Monitoring, then select the Log tab. Any relevant status information is displayed there.

6. Return to the Devices navigation tree to view the refreshed device information

(configured and unconfigured): l l l

To display SCSI and NDMP libraries available to the NetWorker server, select Libraries in the navigation tree. Any available library or silo appears in the Libraries detail table.

To display stand-alone devices available to the NetWorker server, select

Devices in the navigation tree. Any available stand-alone device appears in the Devices detail table, along with devices available in libraries.

To display the libraries and devices that are available to a storage node, select the storage node in the navigation tree. Available storage nodes

SCSI and VTL libraries 143

Backup Storage

144 appear in the table. Double-click a storage node to see its details, along with the devices that are available in the storage node.

Barcode labeling tips

The NetWorker server uses volume labels and barcode labels to identify volumes. Both label types are recorded in the media database. The volume label is also recorded internally on the media (internal volume label). The NetWorker server uses barcode labels to inventory volumes, and uses volume labels to identify the volumes needed for backup and recovery. A requirement to match the volume label with the barcode label can be set in the library’s Properties window.

Follow these guidelines when using barcode labels with the NetWorker software: l l l l l l l l

When NetWorker software relabels volumes automatically, it reuses the original volume label name. A label name can be changed only if the volume is relabeled manually. The NetWorker software scans the barcode label during the labeling process and updates the media database with the new volume name and its associated barcode label.

Do not use identical barcode labels for any of the NetWorker volumes. The use of identical labels defeats the purpose of using barcode labels, which is to facilitate the inventory process and ensure label accuracy.

Volume names must be unique on the NetWorker server. Give each volume a unique volume label. If a second volume is labeled with an existing barcode label and the Match Barcode Labels attribute in the library’s properties is enabled, the

NetWorker server displays an error message and does not allow the second volume to be labeled. The error message identifies the library slots containing the two volumes with identical labels and the barcode label.

To correct this problem, either apply a different label to one of the volumes and restart the labeling process, or disable the Match Barcode Labels attribute in the library’s properties while labeling the second volume.

It is not necessary to label existing volumes with barcode labels if they are stored in a vault or offsite for long periods. These volumes are rarely, if ever, inventoried.

Before using barcode labels on existing volumes, affix the barcode labels to them.

Then, load and mount each volume individually, so that the NetWorker server can match the barcode label with the existing volume label.

Record the volume label on the tape.

A variety of barcode labels can be purchased from third-party vendors. Choose from among numeric labels, alphanumeric labels, or a special combination of numbers and characters. Furthermore, barcode labels can be ordered to match a current volume labeling scheme.

Use a consistent labeling scheme. If volumes are labeled with the server name and an extension such as “001,” order a range of labels starting with “server_name.

001” and ending with “server_name.100”, or as wide a range as necessary.

Instructions for barcode labels should be provided with the library hardware documentation. Contact the hardware manufacturer with questions about barcode labels. A consistent labeling scheme helps better organize and track volumes. It also facilitates the inventory process if all of the volumes, use barcode labels.

Configuring a library to use volumes with barcodes

Barcode labeling tips on page 144 provides more information.

Procedure

1. In the Administration window, click Devices.

EMC NetWorker 9.1 Administration Guide

Backup Storage

2. Open the Libraries folder. The Libraries detail table appears.

3. Right-click the appropriate library, and select Properties. The Properties window appears.

4. Select the Configuration tab.

5. In the Media Management area of the Configuration tab, select: l

Bar Code Reader l

Match Bar Code Labels

6. Click OK.

Using unmatched volume and barcode labels

Note

If unmatched volume and barcode labels are to be used, ensure that labels are attached to the outside of the volumes.

Procedure

1. Apply barcode labels to the volumes.

2. Place the volumes with the barcode labels in the library.

3. In the Administration window, click Devices.

4. Open the Libraries folder. The Libraries detail table appears.

5. Right-click the appropriate library, and select Properties. The Properties window appears.

6. Select the Configuration tab.

7. In the Media Management area of the Configuration tab: l l

Select Bar Code Reader.

Ensure that Match Bar Code Labels is not selected.

8. Click OK. The NetWorker server uses the next available label from the label template for the volume name. It labels the volumes and records both labels in the media database.

9. Inventory the volumes to ensure that the NetWorker server has the most current volume information.

10. Use Media

Volumes to match the correct volume labels to the barcode labels.

Consider making a list of the name correlations.

Note

If the barcode function is enabled, but no barcode label is affixed to the volume, an error message indicates that a barcode label does not exist.

Identifying a specific network interface for device scan operations

If the NetWorker server has multiple network interfaces, you can specify that a specific network interface be used for scan operations. In this case, the dvdetect

SCSI and VTL libraries 145

Backup Storage

(device scan) program will use the specified network address or hostname to communicate with the NetWorker server.

Procedure

1. In the server’s Administration interface, click the Devices button.

2. Select View

Diagnostic Mode.

3. In the left pane, click on the Storage Nodes folder.

4. In the right pane, select a storage node.

5. Right-click the storage node and select Properties.

6. Select the Configuration tab.

7. In the Server network interface field, type the network address or the unique hostname of the network interface on the NetWorker server that is to be used.

8. Click OK.

Media Library parallelism

Use the Max parallelism attribute on the Configuration tab of the Library resource to define the media library parallelism.

Media library parallelism allows you to define the maximum number of available devices for inventory and label operations.

EMC recommends that you set the Max parallelism attribute of the Library resource to one less than the number of devices within the library, which allows you to reserve on device for recovery operations.

To improve the efficiency of library operations that operate on multiple volumes, use multiple devices in parallel for these operations. However, you may wish to restrict the number of devices that NetWorker uses for inventorying and labeling operations, to ensure that some devices are available for other library operations.

Managing the library configuration

This section provides detailed information about managing a tape library in the

NetWorker environment.

Auto Media Management

Auto Media Management gives the NetWorker server automatic control over media that are loaded in the storage device.

When you enable the Auto Media Management feature during device configuration, the NetWorker server automatically: l

Labels the volume (recognizes EDM labels and does not overwrite them).

l l

NOTICE

If the Auto Media Management feature is enabled, the NetWorker server considers volumes that were labeled by a different application to be valid re-label candidates.

Once the NetWorker server re-labels the volume, the previously stored data is lost.

Mounts the volume.

Overwrites volumes that are consider to be unlabeled. The NetWorker server considers a volume to be unlabeled under the following conditions: n

Has no internal label.

146 EMC NetWorker 9.1 Administration Guide

Backup Storage n n

Is labeled with information other than a NetWorker label.

Is labeled with a NetWorker label, but the density that is indicated on the internal label differs from that of the device where the volume is mounted.

Recycles volumes eligible for reuse that are loaded into the device.

When you do not enable the Auto Media Management feature, the NetWorker server ignores unlabeled volumes and does not use the volume for backup.

The Auto Media Management feature can re-label a volume that has a different density, it is possible, inadvertently, to overwrite data that still has value. For this reason, be careful if NetWorker volumes are shared among devices with different densities.

Existing tapes with NetWorker labels

When Auto Media Management is used with tapes that have NetWorker labels that have not been recycled, the volumes must be removed from the media database before a utility such as tar is used to overwrite the labels. Also ensure that the tapes have been fully rewound before overwriting the labels. Auto Media Management can then properly relabel the tapes.

Auto Media Management for stand-alone devices

The Auto Media Management feature can be enabled for stand-alone devices during manual device configuration, or from the Properties window after configuration.

When Auto Media Management is enabled for a stand-alone device, the following processes occur when a volume becomes full during a backup: l l l l l

A notification is sent that indicates that the server or storage node is waiting for a writable volume. Simultaneously, the NetWorker server waits for the full, verified volume to be unmounted.

The device is monitored and the software waits for another volume to be inserted into the device.

After a volume is detected, a check is performed to determine whether the volume is labeled. If so: n n

The volume is mounted into the device.

The NetWorker server checks to see whether the newly mounted volume is a candidate to receive data:

1. If yes, the write operation continues.

2. If no, the NetWorker server continues to wait for a writable volume to continue the backup.

If the volume is recyclable and is a member of the required pool, it is recycled the next time a writable volume is needed.

If the volume is unlabeled, it is labeled when the next writable volume is needed for a save. Note that Auto media management does not label disk type devices such as AFTD and Data Domain.

NOTICE

If a partially full volume is unmounted, the NetWorker server automatically ejects the volume after a few seconds. If a stand-alone device is shared between storage nodes, then Auto Media Management should not be enabled for more than one instance of the device. Enabling Auto Media Management for more than one instance of the stand-alone device will tie up the device indefinitely. No data is sent to the device and no pending message is sent.

SCSI and VTL libraries 147

Backup Storage

148

Enabling Auto Media Management for libraries

Auto Media Management is not enabled for libraries during autoconfiguration. Auto

Media Management for a library can be set by changing the library’s properties after configuration.

Procedure

1. In the server’s Administration window, click Devices.

2. Select the Libraries folder in the navigation tree. The Libraries detail table appears.

3. Right-click the library, and select Properties. The Properties window appears.

4. Select the Configuration tab.

5. In the Media Management area, select Auto Media Management.

6. Click OK.

Labeling volumes

The NetWorker software applies a label template to create a unique internal label for each volume. The label corresponds to a pool and identifies the pool for the volume during backup and other operations.

Several preconfigured label templates are supplied with the NetWorker software. You cannot delete these preconfigured label templates.

Naming label templates

on page 74 provides more information.

When you label a volume, the labeling process: l l l

Writes a label on the volume.

Adds the volume label to the media database.

Prepares tape media to have data written to it.

When you re-label tape, the data on the tape is effectively gone.

During data recovery, the server requests the volume that contains the required data, identifying the required volume by the name with which it was labeled.

Label templates

Several preconfigured label templates are supplied with the NetWorker software.

These preconfigured label templates cannot be deleted.

Naming label templates on

page 74 provides more information about label templates and preconfigured label template.

Labeling or re-labeling library volumes

Labeling volumes in a library is time-consuming, so consider labeling volumes before it is time to back up or recover files. When a volume is re-labeled, that volume is initialized and becomes available for writing again.

Procedure

1. In the Administration window, click Devices.

2. In the left pane, select Libraries.

A list of libraries appears in the right pane.

3. Right-click the library and select Label.

Details for the selected library appear, including divided tables for devices and slots. The Label Library Media dialog box also appears.

4. From the Target Media Pool list, select the pool for the volume.

The pool determines the label template that is used to label the volume.

EMC NetWorker 9.1 Administration Guide

Backup Storage

5. To require manual recycling of the volume, select Allow

Manual Recycle.

With manual recycling, the volume is not automatically marked as recyclable when all save sets expire. You must manually mark the volume as recyclable.

NOTICE

A volume that has been set to manual recycle retains that setting, even after the volume is re-labeled. You must explicitly reset the volume to automatic recycle by right-clicking the volume in the Media window, selecting Recycle, and then selecting the Auto option.

6. To be prompted before the existing label is overwritten, select Prompt to

overwrite label.

7. Click OK.

The Library Operation dialog box appears, stating that the library operation has started.

8. To track the status of the label operation, click Monitoring in the

Administration window.

9. If you selected Prompt to overwrite label, confirm the overwrite of the existing volume label with a new label: a. Right-click the label operation in the Monitoring window and select Supply

Input.

A confirmation message appears.

b. Click Yes.

Verifying the label when a volume is unloaded

If a SCSI reset is issued during a backup, the volume rewinds and NetWorker may overwrite the volume label.

To detect if the label is overwritten in this circumstance, select the Verify label on

eject checkbox in the Device resource, or set the Verify label on unload setting in the

Jukebox resource to Yes. With these settings, NetWorker verifies that a volume label exists before ejecting the volume. If the volume label cannot be read, all save sets on the volume are marked as suspect and the volume is marked as full.

Empty slots in label operations

Slots that have been intentionally left empty (such as bad slots) are skipped during labeling operations. The NetWorker software logs a message similar to: “Slot 5 empty, skipping.”

Barcode labels

The option to label a library volume with a barcode is available during automatic device configuration. This option can be set in the library’s Properties tab after configuration.

Barcode labels make volume inventory fast and efficient. They eliminate the need to mount the volumes in a device. The library scans the external barcode labels with an infrared light while the volumes remain in their slots. Inventorying with barcode labels greatly reduces the time needed to locate a volume or determine the contents of a library.

Barcode labels also provide greater labeling accuracy. The labels are placed on the volumes before the volumes are loaded and scanned in the library. Once the library has scanned the barcode, the NetWorker server records and tracks the label in the media database. The NetWorker server uses barcode labels only to inventory volumes. A volume must have a label, but it need not have a barcode label.

SCSI and VTL libraries 149

Backup Storage

Note

Libraries include hardware that reads barcode labels. The barcode information is then forwarded to the NetWorker server. Problems reading barcode labels indicate hardware problems. In the event of a barcode-related problem, consult the library’s documentation or the hardware vendor.

Requirements for performing an inventory with barcodes

To perform an inventory by using barcodes, the following requirements must be met: l l l

The library must have a barcode reader.

A barcode label must be present on the tape.

The location field within the NetWorker media database must be correct or null. To view the location field, use the mmlocate command.

Device Service mode

Use the service mode setting to take a device offline temporarily. Service mode differs from the disabled state in that the nsrmmd process is not stopped.

While a device is in service mode, save or recover sessions that are either in process or pending are completed. No new sessions are assigned to the device while it is in service mode.

Although a drive in service mode is taken out of the collection of drives that the

NetWorker software can select for automated operations, the drive is available for some manual operations that use the nsrjb or nsrmm command with the -f option.

For more information, refer to the EMC NetWorker Command Reference Guide or the

UNIX man pages.

The device might also go into service mode, rather than become disabled, if consecutive errors occur in excess of the maximum consecutive error count specified for the device. This means that if there are no hardware issues, the tape can be ejected and used in other drives.

Media handling errors

on page 173 provides more information about how to set the maximum consecutive error count.

Note

The drive must be manually reset to Enabled for the NetWorker software to use the device again.

Setting the Service mode for a device

Procedure

1. Open the device’s Properties window.

2. On the General tab, set Status Enabled to Service.

Reconfiguring a library

Use this procedure to reconfigure a tape library.

Before you begin

To reconfigure a library or to add or remove access paths to the devices in a library, use an account with the Configure NetWorker privilege. This includes access paths that allow libraries to be shared.

150 EMC NetWorker 9.1 Administration Guide

Backup Storage

Note

The following procedure does not support adding NDMP devices to a non-NDMP library if both the NDMP server and the NetWorker storage node are on the same host. Instead, use the jbedit command.

Procedure

1. Run Scan for Devices, in case a device path has been added to, or removed from, the library since the latest scan.

2. In the server’s Administration window, click Devices.

3. Select Libraries in the navigation tree. The Libraries detail table appears.

4. In the navigation tree, right-click the entry for the library to be reconfigured, or open the Storage Nodes folder, open the library folder, and then right-click the library entry there.

5. Select Reconfigure Library. The Reconfigure Library window appears. Note that the storage node name and library name cannot be changed in this window.

6. Make appropriate changes in the Configure devices on various storage nodes

using existing drive connectivity area, selecting or clearing checkboxes as necessary, or using the buttons at the right side of the area (Check All, Clear

All, Reset).

Drives that are already configured to be used by the library display check marks in the boxes that are adjacent to their names: l l

Selecting a box adds the drive to the library.

Clearing a box removes the drive from the library.

The Reset button returns the checkboxes to the condition they had when the Reconfigure Library window was opened.

7. Click Start Configuration to reconfigure, or Cancel to leave the window.

8. Run Scan for Devices to refresh the navigation tree and show the reconfiguration results.

Specifying library slots

The available slots feature controls which volumes the NetWorker server uses for backup. The server uses all volumes in a library to perform recoveries, but the volumes that are automatically selected for backups can be controlled by designating a range of available slots in the library.

Perform the following steps to define the available slots in a tape library.

Procedure

1. Ensure that volumes have been placed in all the available slots of the library so that the NetWorker server can continue uninterrupted with an automatic backup.

With two-sided media, the number of available slots is effectively doubled. For example, with 32 optical disks labeled “jupiter.001.a” to “jupiter.032.b,” there are a total of 64 sides, and therefore, there are 64 slots from which to choose.

2. In the server’s NetWorker Administration interface, select View

Diagnostic

Mode from the menu bar.

3. Click Devices.

SCSI and VTL libraries 151

Backup Storage

4. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

5. In either the navigation tree or in the Libraries detail table, right-click the library on which the slots are to be designated, and select Properties.

6. Select the Advanced tab of the Properties window.

7. In the Media Management Area, in the Available slots field, type a range of contiguous slots, then click + to add the range of slots.

For example (assuming that no slots have already been configured), to designate slots 1 through 3 as available, then skip a defective slot 4, and designate slots 5 through 7 as available, type this information in the Available

Slots field: a. Type 1-3, then click + to add these slots.

b. Type 5-7, then click + to add these slots.

c. Click OK. Slot 4 will be skipped when tapes are loaded.

Reset a library

A library must be reset each time the library and the NetWorker software become out of sync. A library reset can be done using either the Administration interface or the command prompt.

Resetting a library in the Administration interface

To reset a library in the Administration interface:

Procedure

1. In the Administration window, click Devices.

2. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

3. Select a library in the navigation tree or double-click a library in the Libraries detail table to open the double-paned Library Operations view.

The library’s drives are listed in the pane on the left in the Device column. The library’s slots are listed in the pane on the right.

4. Right-click a library in the Device column, and select Reset. You are prompted to reset the library.

5. Click Yes. The Library Operation window appears and displays this message:

The library operation has started.

Please see the Monitoring->Operations screen for its status.

152

6. Click OK.

Resetting a library from the command prompt

Use the nsrjb -HE command to reset a library from the command prompt. For example, the library inventory must be correct after adding drives to an SJI-compliant library, such as adding DLT7000 drives to an ETL 7/3500 device.

To make the NetWorker software aware of these new drives, run nsrjb -HE to reset the library. The -E option reinitializes the library’s element status. Some libraries can track whether there is media in a component in the library. This feature is known as an element status capability.

A series of commands exists that allow direct interaction with libraries (sji commands) and tape drives (cdi commands). These commands should only be used by the most

EMC NetWorker 9.1 Administration Guide

Backup Storage knowledgeable of NetWorker users, as the consequences of using them can be unknown. For information about these commands, refer to the EMC NetWorker

Command Reference Guide or the UNIX man pages.

Deleting libraries

The library's devices remain, and can still respond to NetWorker operations (such as monitoring, labeling, deletion, and so on) after the library definition is deleted. A deletion of a library deletes the library, not its devices.

Procedure

1. In the server’s Administration interface, click Devices.

2. Select Libraries in the navigation tree. The Libraries detail table appears.

3. In either the navigation tree or in the Libraries detail table, right-click the entry for the library to be deleted, and select Delete.

4. When prompted, click Yes.

This message appears:

"Are you sure you want to delete this jukebox? If so, please re-attempt deletion within a minute."

5. Click OK to confirm the deletion.

Library notifications

The NetWorker server uses notifications to send messages about NetWorker events.

Several preconfigured notifications, such as the following, provide information about various situations: l l l

Volumes in the library are 90% full

Library needs more volumes to continue

Library has a mechanical problem l l

Library device needs cleaning

Cleaning cartridge needs attention.

The NetWorker software automatically mounts a required volume as long as the volume is loaded in the library. If a recovery operation requires a volume that is not loaded in the library, the Tape mount request 1 notification sends an alert to

Monitoring > Alerts, with a request to do something with a specific volume.

After a library problem is corrected, it might be necessary to mount a volume so the

NetWorker server can continue to back up or recover files.

Refreshing enterprise library views on request

Procedure

1. From the Console window, click Libraries.

2. In the navigation pane, select a server to update, or select the top item in the hierarchy to update library information for all NetWorker servers.

3. Right-click the server, and select Refresh.

SCSI and VTL libraries 153

Backup Storage

Changing the polling interval for enterprise library views

Enterprise library views are updated periodically without user intervention.

Procedure

1. From the Console window, click Setup.

2. From the Setup menu, select System Options.

3. In the Polling Interval for NetWorker Libraries field, type the appropriate time, in hours.

4. Click OK.

Adding and removing media by using the library front panel

Certain media libraries allow for media to be added and removed by using the front panel display. This operation circumvents the NetWorker server's normal procedures for adding and removing volumes and may cause the server information to become out of sync with the library. Normally, you should use the NetWorker server procedures for adding and removing media, rather than the library's front panel display. This is more efficient and guarantees that the server and the library will be in sync.

If it is necessary to use the library's front panel display to add and remove volumes.

Note

When a library is partitioned, the NetWorker software does not become aware of the partitioning. This means that the entire physical library will be disabled, not just one partition.

Procedure

1. In the Properties window for the Library, on the General tab, set Status

Enabled to Service.

Note

Putting the library in service mode will cancel all operations or wait for operations to complete that cannot be canceled, and then put the library into disabled mode.

2. Once the library is in disabled mode, use the library's front panel to add and remove tapes.

3. In the Properties window for the Library, on the General tab, set Status

Enabled to Enabled.

4. Inventory the library. Inventorying library volumes

on page 160 has information about inventorying libraries.

Volume mounting and unmounting

A volume must be mounted before files can be backed up. If no volume is mounted at the start of a backup, an error message appears and requests that a volume be mounted.

154 EMC NetWorker 9.1 Administration Guide

Backup Storage

Mounting or unmounting a volume in a library

Procedure

1. In the Administration window, click Devices.

2. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

3. Select a library in the navigation tree or double-click a library in the Libraries detail table to open the double-paned library operations view. The library’s drives are listed in the Devices column, and its slots are listed in the Slot column.

4. To mount a volume: a. In the Devices column, select the appropriate drive.

b. In the Volume column, right-click a volume to mount, and select Mount.

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays its status.

c. Click OK.

5. To unmount the volume: a. Right-click the device or the volume in the double-paned table view of the library and select Unmount.

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays its status.

6. Click OK.

Unmounting volumes automatically (idle device timeout)

At times, a volume that is mounted in one device might be needed by another device in the same library. For example, data being recovered by one device could span more than one volume, and the required volume could be mounted on another device. To address this need, a value can be defined in the Idle Device Timeout attribute for that particular library.

The Idle Device Timeout attribute specifies the number of minutes a mounted volume can remain idle before it is automatically unmounted from the device and returned to its slot, where it can then be accessed by another device. For libraries, this attribute appears on the Timers tab of a library's Properties. The default value for a library is 10 minutes.

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Open the Libraries folder in the navigation tree.

3. Right-click the appropriate library in the detail table, and select Properties. The

Properties window appears.

SCSI and VTL libraries 155

Backup Storage

156

4. Select the Timers tab.

5. Specify a value in the Idle Device Timeout attribute.

1. You can also override the library’s Idle Device Timeout attribute for a specific device in the library.

To specify the Idle Device Timeout value for a specific device:

6. In the server’s Administration interface, click Devices.

7. Select View

Diagnostic Mode.

8. Select Devices in the navigation tree. The Devices detail table appears.

9. Right-click the device and select Properties.

10. Select the Advanced tab.

11. Specify a value in the Idle Device Timeout attribute.

The default value is 0 (zero) minutes, which means that the device never times out and the tape must be ejected manually. However, when the value of this attribute is set to 0, the value specified in the device library’s Idle Device

Timeout attribute will take precedence.

Mounting or unmounting a volume in a stand-alone tape drive

Procedure

1. Manually insert a volume in the stand-alone drive, or ensure that a volume is already loaded.

In a stand-alone device, a volume that has been loaded into the drive is not considered to be mounted until it has been explicitly mounted in the user interface or from the command prompt.

2. In the Administration window, click Devices.

3. Select Devices in the navigation tree. The Devices detail table appears.

4. Select the device. To mount the volume, in the Devices detail table, right-click the device, and select Mount.

5. To unmount the volume, in the Devices

detail table, right-click the device, and select Unmount.

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays its status.

6. Click OK.

Labeling and mounting a volume in one operation (stand-alone tape drive)

When multiple storage devices are connected to the NetWorker server, the device for labeling must first be selected from the list of available devices. Remember that labeling a volume makes it impossible for the NetWorker server to recover original data from that volume.

Procedure

1. In the Administration window, click Devices.

EMC NetWorker 9.1 Administration Guide

Backup Storage

2. Manually insert an unlabeled or recyclable volume in the NetWorker server storage device, or ensure that a volume of this type is already present for the

NetWorker server to access.

3. Select Devices in the navigation tree. The Devices detail table appears.

4. Right-click the stand-alone device in the detail table, and select Label. The

Label window appears: a. Type a unique label name, or accept the default name that is associated with the selected pool.

If the volume is unlabeled, the NetWorker server assigns the next sequential label from the label template that is associated with the selected pool. If a recyclable volume from the same pool is being re-labeled, then the volume label name and sequence number remain the same. Access to the original data on the volume is destroyed, and the volume becomes available.

b. Select a pool on the Pools menu. The NetWorker server automatically applies the label template that is associated with the Default pool unless a different pool is selected.

c. Select the Manual Recycle attribute if the volume should be manually recycled.

If the Manual Recycle attribute is enabled when the volume is labeled, the volume cannot automatically be marked as recyclable according to the retention policy. When a volume is marked as manual recycle, the NetWorker server disregards the assigned browse and retention policies. Therefore, only an administrator can mark the volume recyclable.

A volume that has been set to manual recycle retains that setting, even after re-labeling. A Manual Recycle policy cannot be changed back to Auto

Recycle by clearing the Manual Recycle checkbox. The volume must be explicitly reset to use auto recycle.

d. The Mount After Labeling attribute is selected by default. The NetWorker server automatically labels the volume, and then mounts the volume into the device.

5. Click OK.

6. If the volume is recyclable, a message warns that the named volume is about to be recycled, and asks whether to continue. Click Yes to re-label and recycle the volume.

7. After a volume is labeled and mounted in a device, the volume is available to receive data. Since the NetWorker label is internal and machine-readable, place an adhesive label on each volume that matches that internal volume label.

Configuring a library to use volumes with barcodes on page 144 provides

information on using barcode labels.

Note

If you are in the process of re-labeling a mounted volume and you choose not to overwrite the existing label, the volume is left in an unmounted state. To use this volume, mount it again.

SCSI and VTL libraries 157

Backup Storage

Labeling volumes without mounting

Volumes can be prelabeled without being mounted.

To label a volume without mounting, follow the same procedures as for labeling and mounting in one operation, but clear the Mount After Labeling attribute in the Label window.

Mounting uninventoried volumes

You can mount volumes that are not included in the library inventory, but are valid

(properly labelled) NetWorker volumes.

Procedure

1. In the Administration window, click Devices.

2. Select View

Diagnostic Mode on the toolbar.

3. Manually insert the volume in an empty library slot.

4. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

5. Select the library in the navigation tree in which the volume was manually inserted, or double-click the same library in the Libraries detail table. The

Libraries detail table changes to the double-paned library operations view. The library’s drives are listed in the Devices column, and its slots are listed in the

Slot column.

6. In the Devices column, right-click the library in which the volume was manually inserted, and select Inventory. The Inventory Library window appears.

7. Type the slot number of the volume in both the First and Last field of the Slot

Range.

8. Select Operation Type: either Slow/Verbose (the default) or Fast/Silent.

When Slow/Verbose is selected, the Supply Input option and icon on the

Operations screen of the Monitoring window can be used to confirm the choice to relabel a volume. The device path appears in the Device field.

When Fast/Silent is selected, the Supply Input option and icon are not available, and relabeling proceeds automatically, without user input. The device path does not appear in the Device field.

Entering user input

on page

55 provides details.

9. Click OK.

l l

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays its status.

The NetWorker software then inventories the specified slot.

10. Mount the inventoried volume.

NOTICE

Unlabeled tapes may not be mounted for inventorying. Unlabeled tapes can only be mounted to be labeled. An attempt to mount an uninventoried volume by using unlabeled media results in an I/O error. The volume will also be ejected.

158 EMC NetWorker 9.1 Administration Guide

Backup Storage

Libraries with volume import and export capability

The NetWorker software supports the use of the SCSI-II import/export feature found in many brands of library. Depending on the library model, this feature is also known as cartridge access port (CAP), mail slot, and loading port. The import/export feature deposits and withdraws (ejects) volumes from slots in the library. This feature enables the operator to deposit and withdraw cartridges without invalidating the device inventory list. Normally, if the operator opens the door to load or unload media, the element status of the autoloader is invalidated, which requires the reinitialization the library. The NetWorker server does not, however, automatically inventory the volume after a deposit and withdrawal.

The reinitialization usually consists of the following: l l l

An inventory of all slots

A reset of the robotic arm

A check to see whether each drive is working

The Deposit attribute causes a library to take the first available volume from the CAP and place it in the first empty library slot. The Eject/Withdraw attribute moves a volume from a slot (never from a drive) to the CAP.

Depositing a volume by using the import/export feature

Use these general instructions when working with a CAP. Specific instructions for working with a CAP can vary, depending on the library manufacturer. For specific instructions, refer to the library’s documentation.

Procedure

1. Ensure that volumes are available in the CAP for deposit.

2. In the Administration window, click Devices.

3. Select Libraries in the navigation tree.

The Libraries detail table appears.

4. Double-click the library in which to deposit the volume.

The Libraries detail table changes to the double-paned library operations view.

5. Right-click either the device or the slot, and select Deposit.

You are prompted to deposit the volume.

6. Click Yes. The Library Operation window displays this message:

The library operation has started.

The Monitoring

Operations screen displays its status.

7. Click OK.

8. Click Monitoring to go to the Monitoring window, and then select the

Operations tab.

9. Right-click the User Input icon for the deposit job and select Supply Input.

You are prompted to load the cartridges into the ports and type Yes to continue.

10. Click Yes.

11. Right-click the User Input icon for the deposit job and select Supply Input again.

You are prompted to continue depositing volumes.

SCSI and VTL libraries 159

Backup Storage

160

12. Click Yes to continue depositing volumes, or No when done.

Withdrawing a volume by using the import/export feature

Note

If the library is partitioned into logical libraries and the import/export slots are shared between the partitions, you must withdraw volumes by using the nsrjb command with the -P option to specify the port or ports from which to withdraw volumes. Refer to the nsrjb man page or the EMC NetWorker Command Reference Guide for more information.

Procedure

1. Ensure that the volume to be withdrawn is in a known slot, and that the CAP has an empty port to hold the withdrawn volume.

2. In the Administration window, click Devices.

3. Select Libraries in the navigation tree. The Libraries detail table appears.

4. Double-click the library from which the volume is to be withdrawn. The

Libraries detail table changes to the double-paned library operations view.

5. Right-click the slot that contains the volume, and select Eject/Withdraw. You are prompted to withdraw the volume.

6. Click Yes.

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays its status.

7. Click OK.

8. Select Monitoring

Log to see the result. A successful Eject/Withdraw operation ends with a Succeeded comment in the Log.

Inventorying library volumes

When the NetWorker software labels the contents of a library, the software registers the location of the volumes in the library slots when it assigns the volume label. This process is called taking inventory. When the volumes in the library are inventoried, the

NetWorker software reads the label of each volume and records its slot number. If the volumes are not moved in the library after they have been labeled, then the

NetWorker server can access the volumes because each volume label is assigned to a specific slot.

If, however, the contents of the library are changed without being labeled, or if volumes are moved into new slots, the NetWorker software must be notified that the library now holds a different set of labeled volumes or that the volumes are in a different order. For example, if the library has more than one magazine, the volumes must be inventoried each time that a magazine is removed, and another one is loaded into the library.

When the volumes in a new magazine are labeled, there is no need to inventory them.

The NetWorker software automatically records the slot number in which each newly labeled volume is located.

The NetWorker software can use barcode labels to speed up the inventory process. If the library supports the use of barcode labels, consider using them if large numbers of

EMC NetWorker 9.1 Administration Guide

Backup Storage volumes, and/or if the library contents change often.

Barcode labels on page 149

provides more information on using barcode labels.

Procedure

1. In the Administration window, click Devices.

2. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

3. Select a library in the navigation tree or double-click a library in the Libraries detail table. The Libraries detail table changes to the double-paned library operations view.

4. Right-click anywhere within the Devices pane, and select Inventory. The

Inventory

Library window appears.

5. Type the numbers of the first and last slots to be inventoried in the Slot Range area.

6. Select Operation Type: either Slow/Verbose (the default) or Fast/Silent.

7. Click OK.

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays its status.

8. Click OK. If the volumes do not have barcode labels, the NetWorker software must mount each volume, read its label, and unmount it. In this case, the inventory process can take some time to complete.

Library maintenance

Periodically clean a storage library to keep it working correctly. The NetWorker server provides automatic cleaning of devices located in libraries. The server does not support automatic cleaning for stand-alone devices. Cleaning is an option set during configuration.

The service mode feature allows a library to be taken offline temporarily for cleaning or other maintenance.

Automatic tape device cleaning

Tape device cleaning is an automated, self-contained operation. It is no longer part of a media-loading operation. Tape device cleaning is automatically triggered if one of these conditions exist: l l

The last time the device was cleaned was a full cleaning interval ago.

The Cleaning Required attribute for the device is set to Yes in one of the following ways: n n

Manually by the user.

Automatically by the NetWorker server, after it receives a “device needs cleaning” notification.

When one of these conditions is met for a device, cleaning begins as soon as the device becomes available. Loaded devices are unloaded before a cleaning operation begins. Loading a cleaning cartridge (with the nsrjb -l cleaning cartridge command) to force a cleaning operation is no longer supported.

SCSI and VTL libraries 161

Backup Storage

162

Selecting a tape device manually for cleaning

NOTICE

Do not enable automated cleaning for silos in the NetWorker software. The automated device cleaning feature cannot be used in a silo, because it depends on fixed slot numbers. For information about how to clean devices in a silo, refer to the silo manufacturer’s software documentation.

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Open the Libraries folder in the navigation tree and select the drive that contains the mounted volume with the block size being checked. The drive’s detail table appears.

3. Right-click the drive in the detail table, and select Properties. The Properties window appears.

4. Select the General tab.

5. Set the Cleaning Required attribute to Yes.

Delaying tape device cleaning

Occasionally it is necessary to set the Cleaning Delay attribute in order to allow a tape device to sleep before attempting to unload a cleaning cartridge.

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Select View

Diagnostic Mode.

3. Open the Libraries folder in the navigation tree.

4. Right-click the appropriate library in the detail table, and select Properties. The

Properties window appears.

5. Select the Timers tab.

6. Select a value in seconds for the Cleaning Delay attribute.

Tape alert

The TapeAlert feature provides, among other things, diagnostic information for devices for which hardware cleaning is enabled.

NetWorker provides the following attributes for tape device cleaning: l l

Cleaning required

Cleaning interval l

Date last cleaned

When the Common Device Interface (CDI) is enabled, TapeAlert attributes provide tape drive status. SCSI Commands must be selected for the CDI attribute on the

Configuration tab of the relevant device’s Properties. If CDI cannot be enabled,

TapeAlert is not supported.

Devices that are capable of TapeAlert perform constant self-diagnostics and communicate the diagnostic information via the nsrmmd program to logs that can be viewed in the Monitoring task.

The following TapeAlert attributes are found in the device’s Properties, on the Volume tab.

EMC NetWorker 9.1 Administration Guide

Backup Storage l l

TapeAlert Critical: Displays critical diagnostic information, such as for media or drive failure, when user intervention is urgent and data is at risk.

TapeAlert Warning: Displays a message when the media or device needs servicing.

TapeAlert Information: Displays status information.

The following table describes the nature of the tape alert levels.

Table 31 Tape alert severity

Severity Explanatory

Critical

Warning

Informative

Urgently requires user intervention

Risks data loss

X X

The messages indicate tape and drive states related to tape drive read/write management, cleaning management, or drive hardware errors.

Informative messages

Informative messages indicate status information: l l

A data or cleaning tape is nearing its end of life.

A tape format that is not supported.

Note

When automatic cleaning is enabled, a diagnostic message to indicate that a drive needs cleaning initiates NetWorker drive cleaning.

Warning messages

Warning messages indicate the following types of drive errors: l l l l

Recoverable read or write errors occurred.

Media is at end of life.

Read-only tape format is in the drive.

Periodic cleaning is required.

Critical messages

Critical messages are warnings that a drive might be disabled and requires immediate attention to avoid data loss: l l l

Unrecoverable read or write errors occurred.

Tape is marked read-only.

Drive require immediate cleaning.

Drive is predicting hardware failure.

Informative and warning messages should clear automatically by nsrmmd once the reported issue is handled.

Critical messages about hardware errors are not cleared by nsrmmd because they might indicate intermittent hardware problems.

SCSI and VTL libraries 163

Backup Storage

Troubleshooting libraries and devices

This section provides detailed information about how to troubleshoot issues with libraries and devices, including how to correct drive ordering issues and block size issues between UNIX and Windows devices.

Troubleshooting autoconfiguration failure

Common symptoms of library autoconfiguration failure include the following: l l

The library is not listed in the Libraries folder in the Administration interface.

The library is listed, but is listed as being unconfigured.

Common causes include: l l

Device drivers are not properly installed.

Autodetection fails to match a detected library with its devices due to: l n

Out-of-date device firmware.

Failure of the library to return its devices’ serial numbers.

Autodetection failed to start on the storage nodes.

Procedure

1. Check Monitoring

Log for relevant messages.

2. From the command prompt, type the following command to verify that the library returns the serial numbers of its devices: sn -a b.t.l.

where b.t.l. refers to the bus target LUN of the library. If the bus target LUN is not known, run the inquire command first, to obtain this information.

Library configuration using the jbedit command

If the autoconfiguration program cannot be used, the jbedit (jukebox edit) program can be used as a fallback means of editing library configurations. This command can be run on a NetWorker server, storage node, or client (if the client is a storage node). It operates without disrupting any backup or recovery operations on the library.

Running the jbedit program requires Configure NetWorker user privileges.

The jbedit program supports all direct-attached SCSI/SJI, SAN, and NDMP libraries.

The jbedit program is not intended to be a full-fledged editor of the Library resource. The editing of Library resource attributes should be done as described in

Reconfiguring a library on page 150. The jbedit options provide selection lists that

make it easy to find drives or devices to be added or deleted.

The following table lists the most commonly used jbedit program options.

Table 32 Common jbedit options

Option

-a

-d

-j

Description

Add a drive or device.

Deletes a drive or device.

Name of the autochanger to be edited.

164 EMC NetWorker 9.1 Administration Guide

Backup Storage

Table 32 Common jbedit options (continued)

Option

-f

-E

Description

Name of the device to be added or deleted.

Element address of the device to be added or deleted.

The EMC NetWorker Command Reference Guide or the UNIX man page provides a detailed description of the jbedit command, its options, and associated diagnostic messages.

Device ordering

The NetWorker server uses logical device names assigned by the operating system when communicating with devices. It is possible for the operating system to reassociate logical device names with the physical addresses of the devices, generally after rebooting the host or after plug-and-play events. This may cause device reordering, where the physical device will have a different device filename. As a result, tape devices configured in the NetWorker software no longer match the names of the devices as recognized by the operating system.

If device reordering occurs, the NetWorker software is unable to use any affected drives until the configuration is manually corrected.

The NetWorker server detects device reordering events by comparing the current serial number of the device to the serial number of the device at configuration. If the serial numbers do not match, the NetWorker server stops all operations on that device and an error message will be posted, similar to the alert identified for device serial

number mismatch in the table Preconfigured notifications

on page 664. CDI must be enabled for this functionality.

Setting the common device interface

on page 172 provides more information about enabling CDI.

Detecting device ordering issues

To determine if there is a problem with device ordering in your environment, you first determine if the device order that appears in nsrjb output matches the device order from the inquire and sjisn commands, then verify that the device configuration within your NetWorker configuration conforms to this.

Procedure

1. Execute the inquire command with the -cl option to determine the device path, scsi address, and serial number of the device.

2. Execute the sjisn command to determine the current order of the devices: sjisn [email protected]

where bus.target.lun is the SCSI address of the robotic arm returned by the inquire command in step 1, for example, 1.2.0.

3. Match the serial numbers of the devices in the sjisn output to the device names that correspond to these serial numbers in the inquire -cl output. This will give you the current device order by device filename.

4. Execute the nsrjb command to determine the order of devices as configured in NetWorker. Drive entries towards the end of the nsrjb output list the device order as configured in NetWorker.

SCSI and VTL libraries 165

Backup Storage

166

5. Compare the device ordering as determined in step 3 and step 4. If the device ordering in these two steps do not match, the device ordering has changed and the library will need to be reconfigured.

Drive ordering change corrections

After a drive ordering change has taken place and the NetWorker software is no longer correctly communicating with devices, you can correct the problem within your

NetWorker configuration by using the NetWorker Console or the jbedit command line program.

Using NetWorker Console to correct drive ordering changes

You can correct drive ordering changes by using the NetWorker Console.

Procedure

1. Ensure that you have a current backup of the resource database.

2. Delete the library resource in the NetWorker Console.

Deleting libraries on page

153 provides details.

3. Rescan the library.

Scanning for libraries and devices

on page 143 provides more information.

Using the jbedit command to correct drive ordering changes

You can correct drive ordering changes by using the jbedit command.

Procedure

1. Use the jbedit command with the -d option to delete devices from the

NetWorker configuration.

2. Use the jbedit command with the -a option to add the devices again.

Library configuration using the jbedit command on page 164, or the UNIX man

page for jbedit or the EMC NetWorker Command Reference Guide provides more information on the jbedit command.

Clearing device ordering/serial mismatch errors from the NetWorker Console

After a device ordering error has been detected, a message is displayed in the Alerts and Notifications windows of the NetWorker Management Console, as well as the log files. The error message is similar to the following:

“Check system device ordering. Moving device on %s to . To correct, scan for devices in NMC and re-enable the device.”

An Event ID for the error is also created, which will be removed along with the alert when the problem is resolved. You can resolve the problem and clear the error message.

Procedure

1. Disable the drive.

2. Perform one of the above procedures to correct the problem.

3. Re-enable the drive, and retry the operation that was being performed prior to receiving the error.

Results

The Alert will be removed and the event dismissed.

Tape drive number reordering (Microsoft Windows only)

If more than one tape drive is attached to the NetWorker server when both the server and drives are shut down, restart all of the tape drives, either before or immediately

EMC NetWorker 9.1 Administration Guide

Backup Storage after the NetWorker server is restarted. If Windows does not locate all of its previously configured tape drives at the time of startup, it automatically reassigns the tape registry name.

For example, assume that these three tape drives are attached to the server: l l

The first one, \\.\Tape0, is a 4 mm tape drive.

The second, \\.\Tape1, is an 8 mm tape drive.

The third, \\.\Tape2, is also an 8 mm tape drive.

If only the second and third tape drives are restarted, Windows reassigns the tape registry numbers so that the second storage device becomes \\.\Tape0 and the third storage device becomes \\.\Tape1. The tape registry numbers no longer match the defined storage devices within the NetWorker software. As a result, the server mishandles the drives and their volumes.

It might be easier to leave a nonoperational drive (device) attached to the server until a replacement is available. If the drive is removed, the name must be deleted, and then the new drive must be added.

To disable the drive, select No for the Enabled attribute in the device’s Properties.

Device calibration

For information about the frequency and method for calibrating the loading mechanism for the device, refer to the library manufacturer’s documentation.

SCSI data block size issues between UNIX and Windows

Different SCSI hardware limitations exist between UNIX and Microsoft Windows operating systems. This can lead to data block size compatibility problems (although they are less likely to occur now than in the past, given larger Fibre-Channel capacities). For example, with a device defined in UNIX that is physically attached to a

Windows HBA, it is possible to define a block size greater than that allowed by the

Windows hardware. This could lead to I/O errors in both write and read states on the device. In order to use both operating systems, it is necessary to determine a block size that is acceptable to both.

NOTICE

In NetWorker 8.0.1 and later, the default block size for an LTO device increases from

128 KB to 256 KB. When NetWorker labels a new or used volume in an LTO device and the Device block size attribute of the device is handler default, the label operation uses a 256 KB block size.

Determining the allowable block size

You can determine the allowable block size by checking the Properties window of a mounted volume while in Diagnostic Mode.

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Select View

Diagnostic Mode.

3. Open the Libraries folder in the navigation tree and select the drive that contains the mounted volume with the block size being checked. The drive’s detail table appears.

4. Right-click the drive in the detail table, and select Properties. The Properties window appears.

5. Select the Volume tab. In the Loaded Volume area, one of the displayed volume attributes is the Volume Block Size.

SCSI and VTL libraries 167

Backup Storage

6. Click OK.

Solving block-size compatibility problems

Note

It is also possible to solve problems with block-size compatibility by changing the block size for an entire device type. The change, however, must be made on each storage node where it is to be available. Once the block size is changed, it affects only those volumes that are labeled after the change. Volumes can be relabeled to use the new block size, but if they contain data that should be saved, be sure to clone the data beforehand to a volume that already uses the new block size.

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Select View

Diagnostic Mode on the menu bar.

3. Open the Libraries folder in the navigation tree and select the drive that contains the mounted volume with the block size being checked. The drive’s detail table appears.

4. Right-click the drive in the detail table, and select Properties. The Properties window appears.

5. Select the Advanced tab. In the Device Configuration area, the currently configured Device Block Size value is displayed.

6. Select the appropriate Device Block Size value.

7. Click OK.

Setting the block size for a device type

Procedure

1. Change the block size: l

On UNIX, change the block size by setting this environment variable to the greatest common value for both systems. For example: setenv NSR_DEV_BLOCK_SIZE_MEDIA_TYPE value where: n n

MEDIA_TYPE is the backup device type available to the NetWorker server (also found in the Media Type attribute on the General tab of the device’s properties). The media type syntax must be all uppercase, with underscores (_) replacing blank spaces and hyphens. Therefore, a device displayed in the NetWorker software as "8mm Mammoth-2" would be listed as:

8MM_MAMMOTH_2 value must be a multiple of 32 KB, with a minimum value of 32 KB.

On Microsoft Windows only, install a later model HBA, or upgrade to drivers that can support up to 128 KB blocks. Windows also accepts the same environment variable format as UNIX to set block size.

2. Restart the NetWorker server in order for changed environment variables to take effect.

168 EMC NetWorker 9.1 Administration Guide

Backup Storage

Device block size for read and write operations

The block size for a volume is defined during the label operation. The label operation uses the value defined in the Device block size attribute for the Device or the value defined by the appropriate block size environment variable.

The block size for both read and write operations uses the block size defined in the volume header during the label operation rather than the device block size.

Block-size mode (UNIX/Linux only)

Ensure that the block size mode for tape devices that are used with NetWorker software is set to variable. Otherwise, data recovery might fail. The procedure for setting the device block size varies depending on the operating system.

The operating system’s documentation provides information about setting the tape device block size in the operating system.

Device parameter settings

Device parameter settings can be modified for the devices the NetWorker software uses in two ways: l l

Individually, through the NetWorker Administration interface.

Globally, for all devices through operating system environment variables. The adjustment of environment variables should only be done by users who know the server environment and performance tuning requirements. For example, an administrator who wants to fine-tune performance by changing a certain setting for all LTO devices on a particular NetWorker server.

The variables (and their equivalent names in the Administration interface) are described in the following sections.

Device setting environment variables

There are several device-related environment variables available to configure devices for the NetWorker software.

Device-related environment variables include the following: l l l l l l

NSR_DEV_BLOCK_SIZE_MEDIA_TYPE

NSR_DEV_TAPE_FILE_SIZE_MEDIA_TYPE

NSR_DEV_LOAD_TIME_MEDIA_TYPE

NSR_DEV_LOAD_POLL_INTERVAL_MEDIA_TYPE

NSR_DEV_LOAD_TRY_LIMIT_MEDIA_TYPE

NSR_DEV_DEFAULT_CAPACITY_MEDIA_TYPE where:

MEDIA_TYPE is the backup device type available to the NetWorker server.

Note

The media type syntax must be all uppercase, with underscores (_) replacing blank spaces and hyphens. For example, a device displayed in the NetWorker software as “8mm Mammoth-2” would be listed as: 8MM_MAMMOTH_2

To determine the media type, right-click the device an select the General tab. The

Media Type attribute contains the media type that should be used in these environment variables.

NSR_DEV_BLOCK_SIZE_MEDIA_TYPE

NSR_DEV_BLOCK_SIZE_MEDIA_TYPE is organized in units of kilobytes. This environment variable will cause NetWorker to override the default block-size setting defined for the tape drive in the operating system. The value set must be a multiple of

SCSI and VTL libraries 169

Backup Storage

32, with a minimum value of 32. Maximums are determined by platform, SCSI driver, and device.

For example:

NSR_DEV_BLOCK_SIZE_4MM_20GB=64

For information about using this environment variable to set block-size compatibility

between UNIX and Microsoft Windows. SCSI data block size issues between UNIX and

Windows

on page 167 provides more information.

NSR_DEV_TAPE_FILE_SIZE_MEDIA_TYPE

NSR_DEV_TAPE_FILE_SIZE_MEDIA_TYPE is organized in units of

NSR_DEV_BLOCK_SIZE_MEDIA_TYPE and is the number of blocks written between filemarks. These filemarks are used to locate a particular spot on the tape during recovery, and more filemarks generally lead to faster positioning. For example:

NSR_DEV_TAPE_FILE_SIZE_TZ89=512

On UNIX and Linux platforms, the NetWorker software writes a filemark by closing and reopening the tape device, which takes one or two seconds. If this value is too small, throughput could be slowed and recoveries may take longer to complete.

On Microsoft Windows platforms, the NetWorker software writes asynchronous filemarks. This setting has a minimal effect on performance.

NSR_DEV_LOAD_TIME_MEDIA_TYPE

NSR_DEV_LOAD_TIME_MEDIA_TYPE is the number of seconds that nsrmmd polls and waits for a drive to become ready after the library inserts a tape into the device.

NSR_DEV_LOAD_POLL_INTERVAL_MEDIA_TYPE is used to set the number of seconds nsrmmd waits between polls during load time.

If the value of NSR_DEV_LOAD_TIME_MEDIA_TYPE is too short, there could be unnecessary load failures. If it is too long, then labeling new tapes takes longer than necessary. The minimum allowable value is 10 seconds. The maximum value is 600 seconds. For example:

NSR_DEV_LOAD_TIME_DTL8000=300

NSR_DEV_LOAD_POLL_INTERVAL_MEDIA_TYPE

NSR_DEV_LOAD_POLL_INTERVAL_MEDIA_TYPE is the number of seconds that nsrmmd waits between each attempt to read a newly inserted tape. The minimum allowable value is 1 second, the maximum value is 30 seconds. For example:

NSR_DEV_LOAD_POLL_INTERVAL_DLT=10

NSR_DEV_LOAD_TRY_LIMIT_MEDIA_TYPE

NSR_DEV_LOAD_TRY_LIMIT_MEDIA_TYPE is the number of times that nsrmmd will attempt to open a drive. The nsrmmd program will poll the drive until the limit set in

NSR_DEV_LOAD_TIME_MEDIA_TYPE is reached. After the limit is reached, it will retry until the NSR_DEV_LOAD_TRY_LIMIT_MEDIA_TYPE is reached. The default value and minimum allowable value is 2, the maximum value is 120.

NSR_DEV_LOAD_TRY_LIMIT_DLT=4

NSR_DEV_DEFAULT_CAPACITY_MEDIA_TYPE

NSR_DEV_DEFAULT_CAPACITY_MEDIA_TYPE is the size of the particular tape used to base the percent full calculation. This variable value has no effect on the actual tape capacity. Any integer value is allowed, with a KB, MB or GB designation to indicate a range of values. Any value less than 200 MB will be overridden by the normal default capacity. There is no obvious maximum, with the only practical limitation being the actual storage size. For example:

NSR_DEV_DEFAULT_CAPACITY_DTL7000=12GB

170 EMC NetWorker 9.1 Administration Guide

Backup Storage

Setting device parameters in the NetWorker Administration interface

You can locate and change the device parameters in the Administration interface.

Procedure

1. In the server’s Administration interface, click Devices.

2. Select View

Diagnostic Mode.

3. Select Devices in the navigation tree. The Devices detail table appears.

4. Double-click the device in the devices table or right-click the device and select

Properties. The Properties window appears, with the General tab selected.

5. Select the Advanced tab. In the Device Configuration area, the device settings are the first fields shown. The following table lists the fields and their corresponding environment variables:

Results

Table 33 Device settings and environment variables

Device setting

Device Block Size

Device File Size

Device Load Time

Device Eject Time

Device Poll Interval

Device Min Load Tries

Device Default Capacity

Device Tape Flags

Corresponding environment variable

NSR_DEV_BLOCK_SIZE_

MEDIA_TYPE

NSR_DEV_TAPE_FILE_SIZE_

MEDIA_TYPE

NSR_DEV_LOAD_TIME_

MEDIA_TYPE

None

NSR_DEV_LOAD_POLL_INTERVAL_

MEDIA_

TYPE

NSR_DEV_LOAD_TRY_LIMIT_ MEDIA_TYPE

NSR_DEV_DEFAULT_CAPACITY_ MEDIA_TY

None

When device parameters are set in this interface, it is not necessary to stop and restart the NetWorker server in order for the settings to take effect.

Setting device environment variables on Windows

Setting environment variables for the NetWorker software differs on Windows and

UNIX operating systems.

Environment variables on Microsoft Windows are set using the Control Panel System applet on the NetWorker server.

Procedure

1. Browse to Control Panel

System and Security

System

Advanced

System Settings.

2. In the General tab click Environment Variables...

3. Click New.

4. Specify the environment variable name and value.

5. Stop and start the NetWorker Backup and Recover Server service in order for the environment variables to take effect.

SCSI and VTL libraries 171

Backup Storage

Setting device environment variables on UNIX

Setting environment variables for the NetWorker software differs on Windows and

UNIX operating systems.

On UNIX and Linux NetWorker sources the /nsr/nsrrc file before starting the

NetWorker processes.

Procedure

1. On the NetWorker server, modify the /nsr/nsrrc file. If this file does not exist, create this file as a Bourne shell script file.

2. Add the environment variables in the following format:

ENV_VAR_NAME = value export ENV_VAR_NAME

3. Stop and start the NetWorker server processes in order for the environment variables to take effect.

Setting the common device interface

The common device interface (CDI) allows the NetWorker server to send commands to tape devices. The CDI feature is not supported within an NDMP environment. CDI support can be set in the NetWorker Administration interface.

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Select View

Diagnostic Mode.

3. Select Devices in the navigation tree. The Devices detail table appears.

4. Double-click a device in the Devices table (or right-click the device and select

Properties). The Properties window appears, with the General tab selected.

5. Select the Advanced tab. In the Device Configuration area, locate the CDI settings: l l l l l

Not Used: Disables the CDI feature and uses standard tape driver calls for tape operations.

SCSI Commands: Sends explicit SCSI commands to tape devices.

When enabled, the CDI feature:

Provides clearer tape status messages.

Informs when a tape is write protected.

Enables Tape Alert, which provides diagnostic information for devices.

Although the CDI feature can be disabled through selecting the Not Used option, it can be time-consuming to disable a large number of devices.

In this situation, access the /nsr/debug directory and create a file named cdidisable. Then restart the NetWorker server. This file does not need any content, it just needs to exist. This disables the use of CDI for that server and all storage nodes controlled by that server.

172 EMC NetWorker 9.1 Administration Guide

Backup Storage

Note

Use of CDI does not change what is written to tape. A tape written with CDI enabled can be read with CDI disabled. Conversely, a tape written with CDI disabled can be read with CDI enabled. The CDI feature enables NetWorker software to collect better diagnostic information and facilitates tape usage when enabled. Only set or disable the CDI feature on the advice of an EMC

Customer Support representative. If tape or SCSI issues occur while the CDI feature is enabled, contact EMC Customer Support.

Media handling errors

The architecture of device drivers can produce media handling errors. The NetWorker software automatically retries a failed operation such as a mount or read of a volume.

The number of times the NetWorker software retries the failed operation depends on the value of the Max Consecutive Errors attribute, which is set in the Advanced tab of the device’s Properties window. The default value is 20. When the device’s Max

Consecutive Errors value is reached, the device stops retrying the operation and becomes disabled.

A mount or read operation might fail for several reasons, for example: l l

Attempts to mount and read a damaged tape in a library can result in a loop of failed actions: the device might repeatedly try to mount the tape, replace it in the slot, and then retry the action with the same result. In this example, to bring the drive back into use, remove the damaged tape, then reenable the device.

A drive that always reports a fixed number of failures before correctly mounting and reading a tape, even if the tape is not damaged, can cause a failure loop. In this example, ensure that the Max Consecutive Errors value is higher than the number of times that particular drive fails before working correctly.

Re-enabling a device

Once the number of retries equals the Max Consecutive Errors value, the device becomes disabled. After the problem that disabled the device has been fixed, the device (drive) must be reenabled before it can be used again.

Procedure

1. When the NetWorker computer is idle, remove any volume from the disabled drive and ensure that the drive is in good working order.

2. In the Administration window, click Devices. The Devices detail table appears.

3. Right-click the drive to be reenabled, and select Properties. The Properties window appears.

4. In the Status area of the General tab, set Enabled to Yes.

5. Click OK.

1. If the disabled drive is part of a library, it might be necessary to reset the device. To do this:

6. From the command prompt, change the path to the directory that contains the

NetWorker binaries.

7. Type this command:

nsrjb -HE

SCSI and VTL libraries 173

Backup Storage

NOTICE

A device retains it enabled or disabled status in the Properties window and in the Devices detail table regardless of whether its storage node is enabled or disabled. Therefore, it is possible that the storage node Properties window is set to disabled while its devices appear to be enabled in the GUI.

Silo libraries

This section describes silos and silo devices. Silos and libraries are managed similarly by NetWorker software.

A silo tape library (STL) is a peripheral that usually contains many storage devices.

Silos libraries have a robotic controller that moves tape media between slots and devices. Silos do not use a SCSI interface to access and control the media movements. Media movements are controlled by a separate host that is called the silo server. The silo server uses silo management software to manage media movement requests over the network. The silo vendor provides the silo management software.

The silo server cannot be the same computer as the NetWorker server.

The silo can be shared among many applications, systems, and platforms. As with libraries, silos make data and media operations more automatic. Silos can load, change, and manage volumes, and clean the devices automatically.

NetWorker only supports silos that use the Automated Cartridge System Library

Software (ACSLS) Manager software.

NetWorker software interactions with a silo

A NetWorker server acts as a client of the silo management software, which resides on the silo server. The NetWorker server communicates with the silo through the Silo

Tape Library Interface (STLI), which must be installed on the NetWorker server that uses the silo.

To access the volumes and devices in a silo, the NetWorker server sends a request to the silo management software, in the form of an STLI call. For example, to mount a volume in a silo device, the NetWorker media service sends a request to the silo management software to mount the volume into a particular device in the silo. The silo server responds to the request and mounts the volume in the requested device.

The silo management software controls many of the operations that NetWorker software controls with a library. For example, the silo management software keeps track of the slot where each silo volume resides, and might control the deposit and withdrawal of volumes, as well as automated cleaning of silo devices.

Naming conventions for silo devices

The silo name of the storage devices is supplied during the configuration process. The silo name is the name that the silo management software uses to refer to the storage device. Depending on the type of silo, the device name can take several forms. This section describes the naming conventions of the currently supported silos.

StorageTek device naming conventions

The StorageTek (STK) silo management software uses either a program that is called

ACSLS that runs on a UNIX system, or a program that is called Library Attach that runs on a Multiple Virtual Storage (MVS) system. These programs name devices according to a coordinate system based on the physical location of the devices in the

174 EMC NetWorker 9.1 Administration Guide

Backup Storage silo. When you configure the silo in NetWorker, you supply the name of the silo that the silo management software uses to refer to the storage device.

For tape drives, the name consists of four digits that are separated by commas: l l l

The first digit refers to the automated cartridge system (ACS) with which the drive is associated.

The second digit refers to the library storage module (LSM) in which the drive is located.

The third and fourth digits refer to the panel and slot location in which the drive is located.

A typical name for an STK drive is similar to: 1,0,1,0.

You cannot determine the drive names from the NetWorker software. Contact the silo administrator for the drive names of the devices that the NetWorker server can use.

To connect to more than one drive, determine the SCSI IDs for each drive and correctly match the IDs to the silo names. If the operating system device names and silo names are accidentally swapped, NetWorker can only mount and unmount volumes. NetWorker cannot read or write to the volumes after they are mounted. To reconfigure the device names correctly, modify the Library resource in the

Administration window and change the order of the device names in the STL Device

Names attribute.

Installing a silo

Procedure

1. Install the silo management software on the silo server.

2. If required, install the STLI library on the NetWorker server. For more information, refer to the documentation from the silo vendor.

For example, for a NetWorker server or storage node running Windows to control an STK silo, the libattach program must be installed.

On UNIX systems, do not install the STLI library because all the necessary software is installed when the NetWorker software is installed.

3. Ensure that the NetWorker server is properly connected to the media devices in the silo.

4. Add the silo. Configuring silo libraries on page 175 provides further details.

Configuring silo libraries

Procedure

1. In the server’s Administration interface, click Devices.

2. Open the Storage Nodes folder in the navigation tree.

3. Right-click the storage node to which the device is to be configured, and select

Configure All Libraries (which is available from many of the menus throughout the Devices task). This action opens a wizard that can configure all detected libraries, except those libraries that are explicitly excluded in the library exclusion list during configuration.

Silo libraries 175

Backup Storage

176

Note

If Configure All Libraries is started from the server folder instead of from the

Storage Node folder, then all storage nodes on the NetWorker server are automatically selected for configuration in the wizard.

The Configure All Libraries wizard appears, and allows the user to step through library configuration, including the following input (some of which is filled in by default): l l l l

Library type (select STL Silo).

Adjust the Enable New Device option, if required.

Current server sharing policy (use maximal sharing with Dynamic Drive

Sharing [DDS]).

Storage nodes on which the libraries should configure. You can select a storage node to see its details that are displayed. If the appropriate storage node is not listed, click Create a New Storage Node. When creating a storage node, replace the default value in the Name field with the name of the new storage node: a. Update storage node properties, if required.

b. Type the Silo Controller count, which sets the number of silos to be configured for the selected storage node. The default is 1. If a silo count of greater than one is selected, then a library name and hostname must be typed for each one.

c. Type the Hostname of the silo controller.

d. (Optional) Use the Test Silo Controller Connectivity button to see whether the connection to a silo controller works. Use it once for each silo. If the connection to a given silo fails, an error message appears.

4. Click Start Configuration after filling in the requested information. The

Configuration window displays a message that the Configure All Libraries process has started, and that the configuration activity can be viewed by checking the Monitoring

Log screen for status.

5. Click Finish on the Configuration window to close the configuration wizard. If problems occur during configuration, then the Back button on the

Configuration window becomes active, which allows the user to return to the input screen to adjust input.

NetWorker software with ACSLS silos

In this section, the term “ACSLS server” refers to the name of the system that is running any one of StorageTek's library manager programs.

The ssi program is used indirectly by the nsrjb program to communicate with an

ACSLS server. The nsrjb program loads libstlstk, which handles the TCP calls to and from the ssi program. The ssi program then handles all of communication to and from the ACSLS server. Starting with ACSLS version 5.3, it is possible to run either a NetWorker server or storage node on the same host that is running ACSLS.

To configure a library, the ssi and mini_el programs must be running on the system on which library configuration is performed. The ssi and mini_el programs are generally run as background processes, and are usually started automatically by the system.

In addition to the ssi and mini_el programs, a shared library file (usually called libstlstk.xxx where xxx is an operating system-dependent extension) is also

EMC NetWorker 9.1 Administration Guide

Backup Storage required. An appropriate version of this library is installed as part of NetWorker installation.

ACSLS silos and firewalls

With ssi version 2.0, communication with the ACSLS server on a specified port number is supported, using the -a command line option. This is part of the STK firewall enhancement. The ACSLS version 7 must be running on the ACSLS server to use this functionality.

The UNIX man pages for these commands, or see the EMC NetWorker Command

Reference Guide provides information on the ssi and mini_el programs.

Releasing a silo device

When a silo device is configured for use with a NetWorker server, it is possible to restrict silo access only to the NetWorker server. These restrictions allow increased availability to the silo for those with full access. These restrictions can be lifted by using the Release Device feature.

Procedure

1. In the Administration window, click Devices.

2. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

3. Select a silo in the navigation tree or double-click a silo in the Libraries detail table to open the double-paned Library Operations view. The silo’s drives are listed in the Device column. The slots are listed in the Slot column.

4. Right-click a silo in the Slot column, and select Release Device. A window appears and asks whether to release devices.

5. Click Yes. The Library Operation window appears and displays this message:

The library operation has started.

Please see the Monitoring->Operations screen for its status.

6. Click OK.

7. Repeat all steps for each device to be released.

Silo device cleaning

Do not enable automated cleaning for silos in the NetWorker software. The automated device cleaning feature depends on fixed slot numbers, so it cannot be used in a silo, which does not have fixed slot numbers. For information about how to clean devices in a silo, refer to the ACSLS silo manufacturer’s software documentation.

Environment variables for StorageTek silos

Environment variables must be set for StorageTek silos. The following table lists the environment variables to set.

Table 34 StorageTek environment variables

Silo model

StorageTek

Environment variables

For UNIX systems: l

CSI_HOSTNAME = name_of_ACSLS_system

Silo libraries 177

Backup Storage

Table 34 StorageTek environment variables (continued)

Silo model Environment variables

l l

The following commands should also be running on the system and can be in the NetWorker startup script:

<binaries_path>/mini_el &

<binaries_path>/ssi &

For Windows systems:

The LibAttach Configurator program is available from

StorageTek. It creates a ssi process, and a link is available to start the mini_el process from

Start

Programs

LibAttach

menu tree.

Once installed and configured, it starts on restart.

Setting environment variables for UNIX systems

Procedure

1. Create a Bourne shell script file named/nsr/nsrrc on the NetWorker server if it does not already exist.

2. Add the variables in this format:

ENV_VAR_NAME = value export ENV_VAR_NAME

3. Stop and start the NetWorker server daemons in order for the environment variables to take effect.

Media management in a silo

More than one software application can use a single silo. Therefore, media management in a silo requires extra operations to prevent the NetWorker software from overwriting volumes used by other programs.

178

Silo slot numbering

In a library, the NetWorker software specifies many functions by slot number. A library has a fixed number of slots, and NetWorker software uses the slot number to refer to a volume’s physical location.

A silo works similarly, but a silo has a variable number of slots, starting at zero when it is first configured, and limited by the silo license purchased. The fundamental identifier of a silo volume is its barcode, or volser (volume serial number). The volser never changes over the life of a particular volume.

When the nsrjb command lists the contents of a silo, it also lists a slot number. Use the slot number to specify which volumes to mount, unmount, label, and inventory.

Volumes are not always assigned the same slot number in the silo. The slot numbers in the silo are assigned dynamically, based on the sorted order of the barcodes that have been allocated. If additional barcodes that fall earlier in the sort sequence are allocated later, then the slot numbers change for all volumes that are later in the sequence.

EMC NetWorker 9.1 Administration Guide

Backup Storage

The nsrjb UNIX man page or the EMC NetWorker Command Reference Guide provide more information.

Silo volume mounting and unmounting

The mount and unmount operations for silos are the same as for library volumes.

Consider the following when mounting and unmounting library volumes: l l

A volume must be mounted before it can be labeled, read, or had data written on it. The robotic mechanism mounts volumes in the devices of a silo.

Volumes must be unmounted before they can be inventoried in a silo or removed from a NetWorker pool.

Volume mounting and unmounting

on page 154 provides more information.

Silo volume labeling

The NetWorker labels for volumes in a silo include both a regular NetWorker volume label (written on the media of the volume) and a silo barcode identifier. The volume label is usually based on the volume pool’s label template. The barcode identifier is written on a physical label on the outside of the volume, which the barcode reader in the silo can scan during inventory.

Labeling volumes on page 148 and

Barcode labels

on page 149 provide instructions on how to label silo volumes.

The use of barcodes with matching barcode labels and NetWorker volume labels, are both available for a silo. The Barcode Reader attribute must be selected, however the

Match Barcode Labels attribute is optional. When both attributes are selected, the internal volume label that NetWorker software writes on the media of each volume will match the barcode label on the outside of the volume. When the labels match, it is easier to track volumes. But the NetWorker software does not require the internal and external labels to match.

With most silo management software, unlabeled volumes can be used. The silo management software assigns a “virtual” barcode label to those volumes. Although volumes can be used without barcodes, it is difficult to maintain integrity, since once the volume has been removed from the silo, the information about the virtual barcode is lost. Any volume without an actual barcode can be reinserted into the silo under a virtual barcode that NetWorker software (or another application) associates with some of the data.

Using silos with volume import and export capability

NetWorker software supports the use of the import/export feature that is found in many brands of silos. Depending on the silo model, this feature is also known as CAP, mail slot, and loading port. The import/export feature deposits and withdraws volumes from slots in the silo.

The import/export feature enables the operator to deposit and withdraw cartridges without invalidating the device inventory list. If the operator opens the door to load or unload volumes, the element status of the autoloader is invalidated, requiring the timeconsuming operation of reinitializing the silo. Note, however, that NetWorker software does not automatically inventory the volume after a deposit.

Either the NetWorker software or the silo management software can be used to control the import/export feature on the supported silos to deposit and withdraw volumes in a silo. But it is often more efficient to use the silo management software, especially to deposit or withdraw many volumes.

If the import/export feature is set to automatic mode, the silo management software inserts volumes automatically and the NetWorker software cannot be used to insert volumes.

Silo libraries 179

Backup Storage

To issue deposit and withdraw commands: l l

To add and deposit volumes, type:

nsrjb -a -T tags -d

To remove and eject/withdraw volumes, type:

nsrjb -x -T tags -w

where tags specifies the tags or barcodes of volumes in a remote silo.

NOTICE

You cannot deposit a volume from the CAP (I/O Port) using the nsrjb -d command. A silo volume deposit requires the -T and -a options in sequence to add a volume in the media database.

The sequence of operations is: n n n nsrjb -d -T Barcode

Ignore the error message that appears.

nsrjb -a -T Barcode

180

Barcode IDs

A list of available barcode-labeled volumes is available from the silo management software. Refer to the silo manufacturer’s documentation for how to generate the list of barcode IDs.

To specify a barcode identifier or template for the volumes from a command prompt, use the -T option with the nsrjb command. The UNIX man page and the EMC

NetWorker Command Reference Guide provides detailed information about the nsrjb command.

Silo volume allocation

When volumes are added, the NetWorker server is directed to the volumes it can use.

NOTICE

Because silos can be used by more than one software application, it is possible that a different application could read or write to volumes that belong to the NetWorker software. To prevent this from happening, most silo management software includes methods to limit access to volumes based on the hostname of the computer on which various programs run. The NetWorker software does not provide a method for setting up this sort of protection. The silo management software must configure it.

The addition of a volume causes the NetWorker software to query the silo management software to verify that the requested volume exists.

If the volume exists, the volume is allocated to the NetWorker software.

Adding a silo volume

Procedure

1. In the Administration window, click Devices.

2. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

3. Double-click a silo in the Libraries detail table to open the double-paned library operations view. The silo’s drives are listed in the Device column, and its slots are listed in the Slot column.

4. Right-click a silo in the Device column, and select Add. The Add Library

Volumes window appears, with the option to select either Template or List for barcode selection.

EMC NetWorker 9.1 Administration Guide

Backup Storage

5. Select either Template or List to enter barcode volume identifiers.

The Template option allows the use of wildcards in creating a list of barcode

IDs. Each entry should be on a separate line. For example, to name four tapes A01B, A02B, A03B, and A04B, type: l

1-4

The List option allows the entry of barcode IDs, separately. Each entry should be on a separate line. For example, type the name for each tape:

A01B

A02B

A03B

A04B

6. Type the appropriate volume identifiers in the Barcodes field.

7. Click OK (or Cancel, to continue adding to the list).

l l l

Click "+" to add an entry.

Click "<-" to insert above a highlighted selection.

Click "-" to delete an entry.

The Library Operation window displays this message:

The library operation has started.

The Monitoring

Operations > screen displays the status.

8. Click OK.

The Library detail table displays the added volumes.

Inventory silos

Taking inventory of the volumes in a silo ensures that the mapping between slot number and volume name is correct, or reconciles the actual volumes in a silo with the volumes listed in the NetWorker media database.

The slot number of a silo volume is not a numbered slot inside the silo, as it is in a library. The slot number of a silo volume is the number of the volume’s position in the list of volumes in a silo.

The tasks for inventorying volumes in a silo are the same as those for a library.

Inventorying library volumes

on page 160 provides information about inventorying a library.

The NetWorker software examines all of the volumes in the silo and compares the new list of volumes to the NetWorker media database. Then the NetWorker software produces a message listing any volumes located in the silo that are not in the media database.

When the NetWorker software inventories a silo, the silo’s barcode label reader reads the barcode labels on the outside of each volume. When a barcode matches an entry in the NetWorker media database, the volume does not need to be loaded. The inventory proceeds rapidly. If, however, the NetWorker software reads a barcode that does not match any of the entries in the media database, the volume must be mounted and read in order for a proper inventory to be taken.

Troubleshooting a silo

If the particular silo model does not automatically deposit the volume, then place the volumes in the insert area, right-click the volume, and select Deposit.

To perform the Deposit and Add operations from a command prompt:

Silo libraries 181

Backup Storage l l

On silos that require manual depositing, type

nsrjb -a -T tags -d

On silos where the silo management software deposits volumes automatically, such as StorageTek silos, type

nsrjb -a -T tags

where: n n tags specifies the tags or barcodes of volumes in a remote silo.

-d performs the manual deposit.

NetWorker software interactions with a silo on page 174 provides more information on

STLIs.

Deallocating (removing) silo volumes

When an STL volume in a silo is no longer needed, the volume can be deallocated from the silo. Deallocation is basically the same operation as removing a volume from a library. Although the volume cannot be loaded by the robotic mechanism, the entries in the NetWorker media database remain intact. If the volume is allocated again,

NetWorker software can retrieve the data from it later.

Use deallocation when the silo license limits the number of usable slots, or when data is moved offsite for safer storage. When the license limits the number of slots, it might be possible to leave the volumes in the silo, if it is certain that the volumes will not be used by another application. That way, the volumes can easily be added again when the data on them must be accessible.

The allocation operation is not automatic. The volumes must be manually allocated again and reinventoried to let the NetWorker server access the data. If the volume is to be removed from the silo for offsite storage, it must be removed with NetWorker software and then ejected from the silo by using the silo management software.

Procedure

1. Unmount the volume from the device.

Volume mounting and unmounting

on page 154 provides instructions on unmounting volumes.

2. In the Administration window, click Devices.

3. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

4. Double-click a silo in the Libraries detail table to open the double-paned library operations view. The silo’s drives are listed in the Device column.

5. Right-click a silo in the Device column, and select Remove.

The Remove Library Volumes window appears, with the option to select either

Template or List for barcode selection.

6. Select either Template or List to enter barcode volume identifiers.

l l

The Template option allows the use of wildcards in creating a list of barcode

IDs. For example, to name four tapes A01B, A02B, A03B, and A04B, type A0,

1-4, and B.

The List option allows the entry of barcode IDs, separately. For example, type the name for each tape: A01B, A02B, A03B, and A04B.

7. Type the appropriate volume identifiers in the Barcodes field.

8. Click OK.

182 EMC NetWorker 9.1 Administration Guide

Backup Storage l

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays the silo’s status.

9. Click OK. Notice that on return to the Libraries detail table, the removed volumes are no longer listed.

Results

NetWorker software interactions with a silo on page 174 provides information on STLs.

NDMP libraries

NDMP libraries or devices are accessed by using the NDMP protocol and are typically used by network attached storage (NAS) systems. These devices do not allow direct access to control from the host operating system. Control and data movement is performed over the network by using the NDMP protocol.

The NDMP guide provides more information.

NetWorker hosts with shared libraries

The NetWorker software permits different NetWorker hosts (a NetWorker server or storage node) within a datazone to control individual devices within a library. This is known as library sharing.

The presence of a SAN within the datazone is not required for library sharing.

Dynamic Drive Sharing (DDS) does not support sharing libraries across datazones.

How library sharing works

Library sharing enables one NetWorker host to control the library’s robotic arm, while other NetWorker hosts (as well as the host controlling the robotic arm) can each control and use specific library devices. A specific device can be controlled only by a single NetWorker host. The following figure shows how multiple NetWorker hosts can share library devices.

Figure 11 How library sharing works

NDMP libraries 183

Backup Storage

Library task inactivity periods

Library resources include attributes that are used by older, slower libraries that specify the number of seconds a library is inactive after certain operations (such as loading, unloading, or ejecting a volume). For example, once a tape is loaded, the library must read and, possibly, reposition the tape before the next operation can begin. This period of delay is known as sleeping.

While sleeping, the library cannot receive or perform other operations. Without the sleep period, the loading or unloading of volumes might fail.

The NetWorker software automatically configures default sleep periods. Change these values only when troubleshooting a library’s performance, or if a NetWorker technical support specialist requests it. Typically, the higher the sleep values specified in the attributes, the longer it takes the library to perform the task. Be cautious when changing these values.

The sleep attributes and their default values are shown in this table.

Table 35 Library resource sleep attributes

Attribute

Load Sleep

Unload Sleep

Eject Sleep

Deposit Timeout

Withdraw Timeout

Cleaning Delay

Idle Device Timeout

Description Default value

Number of seconds that the

NetWorker software waits for a library to complete loading a cartridge.

15 seconds

Number of seconds that the

NetWorker software waits for a library to complete unloading a cartridge.

60 seconds

Number of seconds that the

NetWorker software waits for an eject operation to complete.

60 seconds

Number of seconds for a library to wait for a tape to be deposited in the mail slot before it times out.

Number of seconds for a library to wait for a tape to be withdrawn from the mail slot before it times out.

Number of seconds that the

NetWorker software waits between the completion of a drive cleaning operation and the ejection of the cleaning cartridge from the drive.

The number of minutes

NetWorker allows a device with a volume to be idle before automatically unmounting it. For specific

15 seconds

60 seconds

10 minutes

184 EMC NetWorker 9.1 Administration Guide

Backup Storage

Table 35 Library resource sleep attributes (continued)

Attribute

Port Polling Period

Description Default value

devices, this value can be overridden.

Unmounting volumes automatically (idle device timeout)

on page 155 provides more information.

Number of seconds for a library to wait before polling a mail slot to check for the updated status.

3 seconds

Server Network Interface attribute

The Server Network Interface attributes in the Device resource are used to determine the network address or the hostname used by the nsrmmd program to communicate with the NetWorker server. Similarly, the Server Network Interface attribute in the

Library resource is used to determine the network address or the hostname used by the nsrlcpd program to communicate with the NetWorker server. These attributes are displayed in the NetWorker Console in diagnostic mode only. The Server Network

Interface attributes are only relevant if the device or library is connected to a storage node.

Note

For devices, the nsrmmd program will read the Server Network Interface value for the first enabled device from the list of storage node devices, and each subsequent nsrmmd started by the NetWorker server will use the same value. Therefore, the

NetWorker server will always use the same Server Network Interface value for every nsrmmd it starts or restarts, regardless of whether or not the Server Network

Interface attribute is different for each device.

Dynamic drive sharing

Dynamic Drive Sharing (DDS) is a feature that provides NetWorker software with the ability to recognize shared physical tape drives. DDS enables NetWorker software to perform the following operations: l l

Skip the shared tape drives that are in use.

Route the backups or recoveries to other available shared tape drives.

Introduction to DDS

DDS controls application requests for tape media and allows the NetWorker server and all storage nodes to access and share all attached devices.

A system administrator can configure DDS by setting a sharing policy for devices that are accessible from multiple storage nodes.

There are two terms that are central to the use of DDS are drive and device. Within the context of DDS, these terms are defined as follows: l

Drive—The physical backup object, such as a tape drive, disk, or file.

Dynamic drive sharing 185

Backup Storage l

Device—The access path to the physical drive.

Note

NetWorker only supports DDS in a storage area network (SAN) Fibre Channel environment and not in a direct-connect SCSI environment.

Benefits of DDS

Enabling DDS on a NetWorker system provides these benefits: l l l l

Reduces storage costs—You can share a single tape drive among several storage nodes. In fact, since NetWorker software uses the same open tape format for

UNIX, Windows, NetWare and Linux, you can share the same tape between different platforms (assuming that respective save sets belong to the same pool).

Reduces LAN traffic—You can configure clients as SAN storage nodes that can send save sets over the SAN to shared drives.

Provides fault tolerance—Within a SAN environment, you can configure hardware to eliminate a single point of failure.

Provides configuration over a greater distance—You can configure a system over a greater distance than with SCSI connections.

DDS configuration overview

The following figure illustrates the DDS process and potential device sharing configurations. This basic configuration consists of a server, two storage nodes, and a library with two tape drives.

Figure 12 Dynamic Drive Sharing

186

In this figure: l

Storage nodes sn_1 and sn_2 are attached to the library.

l l

Each storage node, on its own, has access to drive_1 and drive_2.

With DDS enabled, both storage nodes have access to both drives and can recognize when a shared drive is in use.

EMC NetWorker 9.1 Administration Guide

Backup Storage

This configuration requires two DDS licenses, one for each drive.

Note

Ensure that all applicable devices can be seen from each storage node by running the inquire -l command locally on each storage node.

DDS block-size compatibility between UNIX and Windows

With DDS enabled, drives can be shared between storage nodes on different platforms, such as UNIX and Microsoft Windows. For NetWorker software operations

(such as backups and recoveries) to take place successfully, ensure that the block size is compatible between different platforms or hardware.

To ensure compatibility, make sure one of the following conditions is met: l l

The various storage nodes sharing a drive support the same block sizes.

When a tape is labeled on a drive, it is labeled with the block size defined on the storage nodes.

Block-size incompatibility between UNIX and Windows

Incompatible block-size settings between UNIX and Microsoft Windows storage nodes could result in any of these error scenarios: l l l

A backup taken on a UNIX node might not be recoverable on a Microsoft Windows node if the Windows node does not support large block sizes.

A UNIX process labels and saves data to a tape and leaves the tape mounted. A

Microsoft Windows process subsequently attempts to verify the label on this tape and fails because the label verification is done by reading a header from the data portion.

A tape on a UNIX node is labeled with a large block size. The backup is started on a

Microsoft Windows node and the Windows node attempts to write the backup by using the default block size. Internally, the backup on Windows is written by breaking down the big buffer of data into smaller segments of writable block sizes.

Attempting to recover a specific file on Windows in this situation fails due to positioning errors on the tape. The data is still recoverable from the Windows side, since the NetWorker software will switch from using file and block positioning to reading the tape from the beginning to reach the correct position. The data might not, however, be recoverable from the UNIX side.

Unintended Access to DDS device prevention

The Reserve/Release attribute has been added to the Device resource for tape devices to support Reserve/Release, including the Persistent Reserve commands.

Reserve/Release is a mechanism that uses SCSI commands to attempt to prevent unintended access to tape drives that are connected by using a shared-access technology, such as Fibre Channel, iSCSI, or SCSI multiplexers. It is a “cooperative” and host-based mechanism, which means that all applications should respect the reservations and not purposely break them. Access is granted based on the host system that reserved the device. Other applications that run on that host cannot be prevented from accessing a reserved device.

Reserve/Release cannot prevent a malicious or badly behaved application from accessing a reserved device. It also cannot prevent all problems caused by hardware issues (such as SCSI resets or FC LIPs) from interrupting data access.

The basic sequence requires that a host reserve a tape drive (using specific SCSI commands) before attempting to access the tape drive. If this “reservation”

Dynamic drive sharing 187

Backup Storage succeeds, then the host can use the drive. If the reservation fails (usually because the device is reserved by someone else), then the host attempting the reservation should not attempt to use the drive. When a host has finished using a reserved drive, that host must release the drive by using the appropriate SCSI commands.

The reservation is maintained by the drive itself. With older (called “Simple” in

NetWorker software) Reserve/Release, the reservation is based on the SCSI ID of the system that issued the reserve command. For tape drives connected to Fibre Channel

(FC) using FC-SCSI bridges, the mapping between FC host and reservation is done inside the bridge, since the initiator on the SCSI side is always the bridge itself, regardless which host actually issued the reserve command.

For Persistent Reserve, the reservation is associated with a 64-bit “key” that is registered by the host. Several keys can be registered with a given drive at any given time, but only one may hold the active reservation. NetWorker software uses the

“exclusive” reservation method for Persistent Reserve. Only the host that holds the active reservation is allowed to access the drive.

The Reserve/Release attribute does not support file type or advanced file type devices.

The settings that relate to Reserve/Release and Persistent Reserve are found in a device’s Properties window, on the Advanced tab. They are visible only when diagnostic mode is turned on.

The default setting for Reserve/Release is None. Once any other Reserve/Release setting is selected, it works automatically, without further user intervention. The

Reserve/Release attribute is supported only on Common Device Interface (CDI) platforms, so if the CDI attribute in a device’s Properties is set to Not Used, then

Reserve/Release settings are ignored.

For newer hardware, once a Reserve/Release setting (other than None) has been selected, the appropriate Persistent Reserve commands are automatically issued before a device is opened for reading or writing, and before the device is closed. With older hardware, a SCSI-2 Reserve command is issued before opening the device, and a

SCSI-2 Release command is issued after the device is closed.

Reserve/Release has these possible settings: l l l

None (the default)

Simple

Persistent Reserve l

Persistent Reserve + APTPL (Activate Persist Through Power Loss)

The Persistent Reserve Key attribute has also been added. It is used with Persistent

Reservation calls.

Restrictions for use of the SCSI Reserve/Release setting

There are restrictions for using the SCSI Reserve or Release setting.

Consider the following: l l l

It is available on CDI platforms only. Consequently, since CDI is not supported within an NDMP environment, Reserve/Release is not supported with NDMP.

Not all drives support persistent Reserve/Release. (All drives support at least simple reserve release. The code automatically drops back from Persistent

+APTPL or Persistent to Simple on drives that do not support Persistent.)

SCSI resets can clear Simple reservations at the device.

188 EMC NetWorker 9.1 Administration Guide

Backup Storage l l l

Even with Reserve/Release, there is no guarantee against data loss.

If the operating system has its own Reserve/Release feature, that feature must be disabled in order for the NetWorker Reserve/Release feature to work.

Even if all of the enterprise’s NetWorker storage nodes have this feature enabled, then it is possible that, on the storage node where a backup operation is run, data loss can be caused by the operating system’s utilities or by third-party programs.

DDS attributes in the device properties

Configure the attributes that DDS uses, in the Properties window for a device.

The attributes include: l l

Hardware ID

Shared Devices l l

Hardware ID attribute

The Hardware ID attribute tracks the drives that are shared between multiple hosts.

Device instances that share the same physical drive across multiple hosts have the same hardware ID. The device autoconfiguration process automatically assigns the

Hardware ID to a device, or it is added when manually configuring a device. Users cannot edit the Hardware ID.

You can view the Hardware ID in the Properties window for a device, on the General tab, in the Device Sharing area.

NetWorker generates the Hardware ID when a device is scanned or configured. The

Hardware ID consists of the following components: l l

Hardware serial number

Device type

Worldwide part number (WWPN)

Worldwide name (WWN)

Shared Devices attribute

The Shared Devices attribute appears on the Operations tab of a device’s Properties window when in diagnostic mode. It features values that can be used to manipulate all shared instances of a drive simultaneously. This attribute enables or disables all devices that share the same Hardware ID with a single action. The following table lists allowed values and descriptions for the attribute.

Table 36 Shared Devices attributes

Value

Enable All

Disable All

Done

Description

When selected, enables all devices with the same Hardware ID.

When selected, disables all the devices with the same Hardware ID.

This value is the default setting. After the server has enabled or disabled all devices with the same Hardware ID, the attribute value is reset to Done.

Dynamic drive sharing 189

Backup Storage

You cannot configure the Shared Devices attribute with the jbconfig program.

Idle Device Timeout attribute and DDS

A tape might remain mounted in a drive after a backup completes. Other requests for the drive from another device path must wait during this timeout period. Use the Idle

Device Timeout attribute to adjust the timeout value.

The Idle Device Timeout attribute is not specifically a DDS attribute, but is useful in configuring shared drives. This attribute appears on the device Properties window on the Advanced tab when displayed in Diagnostic Mode. The default value is 0 (zero) minutes, which means that the device never times out and you must manually eject the tape.

If the device belongs to a library, you can also specify the Idle Device Timeout value for all devices in the library. However, the library value will take effect only on those devices whose Idle Device Timeout value is 0. The Idle Device Timeout value for a library is located on the Timer tab of the library Properties window.

Max active devices

In a DDS environment, use the Max active devices attribute, on the General tab of the

Storage Node resource to define the maximum number of active devices for a storage node.

This attribute sets the maximum number of devices that NetWorker may use from the storage node in a DDS configuration. In large environments with media libraries that have a large number of devices, storage nodes might not have the ability to optimize all the drives in the library. The Max active devices attribute allows you to limit the number of devices that the storage node uses at a specified time, which allows the storage node to have access to all the devices in the library, but does not limit the storage node to the number of devices it can fully optimize.

File type devices

File type devices (FTDs) are legacy devices and their use is limited. Continued support for legacy and test purposes is maintained, however you are encouraged to use AFTD or DD Boost devices in preference to FTD. An FTD can be configured on the

NetWorker server by creating a new Device resource in the same manner as for other storage devices.

The following conditions and restrictions apply to FTDs: l

The upper limit of save set size on an FTD may be either: l l l l n n

The upper limits supported by the operating system

The file size specified by the disk device vendor

If multiple FTDs are configured on a system, each device must have a unique name.

To use multiple FTDs on the same disk, partition the disk and create only one FTD per partition.

Dynamic Drive Sharing is not supported.

For FTDs created on a UNIX or Linux network file system (NFS): n n n

The file system used for the FTD must not be used for any other data.

There must be one FTD per NFS system.

The Volume Default Capacity attribute for the FTD must be set to a size that is less than 100 percent of the total capacity of the file system.

190 EMC NetWorker 9.1 Administration Guide

Backup Storage

NOTICE

Data loss will result if a full FTD is made appendable while a backup is pending completion and a save set is partially written to the full FTD. In this case, the partial save set (currently in “incomplete” state) will be overwritten.

FTD capacity issues

For FTDs, the Volume Default Capacity is a hard limit on the amount of data that can be written to the device. The Volume Default Capacity value is an estimate of what the volume capacity is likely to be. If the value is not set correctly, the NetWorker percent-used calculation will be incorrect.

Note

By contrast, AFTDs ignore the Volume Default Capacity value to allow dynamic expansion of disk space.

The Volume Default Capacity attribute displays on the Configuration tab of the Device properties when Diagnostic Mode (View > Diagnostic Mode) is enabled: l l l

To avoid accidentally filling an FTD, set the Volume Default Capacity attribute to restrict the size of the device. For example, if a capacity of 100 MB is set, then the device will be marked full when 100 MB is reached.

Volume Default Capacity attribute must not be set to a value of more than 4 TB.

If the Volume Default Capacity of a volume changes, the changes do not take effect until the FTD is re-created, the directory contents are deleted, and the volume is relabeled.

NOTICE

If the FTD is used before the Volume Default Capacity attribute is set, then the legacy data on that FTD must be staged or cloned to another device. Otherwise, this data will be overwritten.

Full FTD prevention

To prevent the file system from becoming full when backing up data to FTDs, policies can be used to move the data off the disk as soon as necessary. Save sets from FTDs can be staged or cloned to an AFTD to take advantage of advanced file type device features.

To make space for additional backups: l l

Configure a save set staging policy. Staging save sets on page 444 provides

details.

Review and, if required, modify the retention policy of the save sets.

Stand-alone devices

A Device resource must be created for each stand-alone tape device on a storage node. Stand-alone drives must be configured individually.

Storage nodes must have been created before devices can be configured to be used

by them. Storage nodes on page 93 provides information about storage nodes and

how to create them. Note that all scanning for devices is done at the storage node level, and can be done across multiple storage nodes. Only devices that have serial

FTD capacity issues 191

Backup Storage numbers can be autoconfigured. Use the jbconfig command to configure devices that do not have serial numbers.

Note

Devices must be updated to the most recent firmware and drivers.

Autodetecting and configuring a stand-alone tape drive

You can configure a new stand-alone tape drive, automatically by using Scan for

Devices.

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Right-click Devices in the navigation tree, and select Scan for Devices to detect available devices. The Scan for Devices window appears.

3. Click Start Scan.

4. Check the scan status by clicking the Monitoring button and selecting the Log tab. Then return to the Devices navigation tree.

5. Select either the Devices folder or the Storage Nodes folder in the navigation tree. All detected drives are listed. Any still-unconfigured drives are preceded by a circular icon that displays a wrench.

6. Right-click the stand-alone drive to be configured, and select Configure Drive.

A Configuration dialog box appears.

7. Click Yes to confirm that the drive should be configured. The new drive is automatically configured.

Adding a stand-alone device manually

Procedure

1. In the server’s NetWorker Administration interface, click Devices.

2. Right-click Devices in the navigation tree, and select New. The Create Device window appears, with the General tab selected, and a default device path in the

Name field of the Identity area of the window.

3. Replace the default name with the path and name of the device: a. If the device is configured on the server’s storage node, the name is the simple device path, such as /tmp/d0 for a file type device. A tape device on

Windows would have a format similar to \\.\Tape0.

b. If the device is configured on a remote storage node, then the name must indicate that the storage node is remote by including rd= and the name of the remote storage node in the device path. For example, if the remote storage node is neptune, then the device path might be rd=neptune:/tmp/d0 or rd=neptune:\\.\Tape0.

File type devices on page 190 provides instructions and restrictions on

backing up to a file type device.

4. In the Identity area, configure the following: a. In the Comment field, add an optional, descriptive comment.

b. In the Media Type field, select a media type.

192 EMC NetWorker 9.1 Administration Guide

Backup Storage

5. In the Status area, configure the applicable checkboxes: l l

Read Only

Auto Media Management

6. In the Cleaning area, configure the applicable fields: l l

Cleaning Required

Cleaning Interval

The Date Last Cleaned is filled in automatically once a drive has been cleaned.

7. Select the Configuration tab to set attributes, such as: l l l

Target Sessions

Max Sessions

Local Backup to a dedicated storage node

NDMP settings (NDMP remote username and password are required for an

NDMP device that acts as a storage node.)

8. Click OK when the configuration is complete.

Auto Media Management for stand-alone devices

The Auto Media Management feature can be enabled for stand-alone devices during manual device configuration, or from the Properties window after configuration.

When Auto Media Management is enabled for a stand-alone device, the following processes occur when a volume becomes full during a backup: l l l

A notification is sent that indicates that the server or storage node is waiting for a writable volume. Simultaneously, the NetWorker server waits for the full, verified volume to be unmounted.

The device is monitored and the software waits for another volume to be inserted into the device.

After a volume is detected, a check is performed to determine whether the volume is labeled. If so: l l n n

The volume is mounted into the device.

The NetWorker server checks to see whether the newly mounted volume is a candidate to receive data:

1. If yes, the write operation continues.

2. If no, the NetWorker server continues to wait for a writable volume to continue the backup.

If the volume is recyclable and is a member of the required pool, it is recycled the next time a writable volume is needed.

If the volume is unlabeled, it is labeled when the next writable volume is needed for a save. Note that Auto media management does not label disk type devices such as AFTD and Data Domain.

Auto Media Management for stand-alone devices 193

Backup Storage

NOTICE

Mounting or unmounting a volume in a stand-alone tape drive

Procedure

1. Manually insert a volume in the stand-alone drive, or ensure that a volume is already loaded.

In a stand-alone device, a volume that has been loaded into the drive is not considered to be mounted until it has been explicitly mounted in the user interface or from the command prompt.

2. In the Administration window, click Devices.

3. Select Devices in the navigation tree. The Devices detail table appears.

4. Select the device. To mount the volume, in the Devices detail table, right-click the device, and select Mount.

5. To unmount the volume, in the Devices

detail table, right-click the device, and select Unmount.

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays its status.

6. Click OK.

Labeling and mounting a volume in one operation (stand-alone tape drive)

Procedure

1. In the Administration window, click Devices.

2. Manually insert an unlabeled or recyclable volume in the NetWorker server storage device, or ensure that a volume of this type is already present for the

NetWorker server to access.

3. Select Devices in the navigation tree. The Devices detail table appears.

4. Right-click the stand-alone device in the detail table, and select Label. The

Label window appears: a. Type a unique label name, or accept the default name that is associated with the selected pool.

If the volume is unlabeled, the NetWorker server assigns the next sequential label from the label template that is associated with the selected pool. If a

194 EMC NetWorker 9.1 Administration Guide

Backup Storage recyclable volume from the same pool is being re-labeled, then the volume label name and sequence number remain the same. Access to the original data on the volume is destroyed, and the volume becomes available.

b. Select a pool on the Pools menu. The NetWorker server automatically applies the label template that is associated with the Default pool unless a different pool is selected.

c. Select the Manual Recycle attribute if the volume should be manually recycled.

A volume that has been set to manual recycle retains that setting, even after re-labeling. A Manual Recycle policy cannot be changed back to Auto

Recycle by clearing the Manual Recycle checkbox. The volume must be explicitly reset to use auto recycle.

d. The Mount After Labeling attribute is selected by default. The NetWorker server automatically labels the volume, and then mounts the volume into the device.

5. Click OK.

6. If the volume is recyclable, a message warns that the named volume is about to be recycled, and asks whether to continue. Click Yes to re-label and recycle the volume.

Configuring a library to use volumes with barcodes on page 144 provides

information on using barcode labels.

Note

If you are in the process of re-labeling a mounted volume and you choose not to overwrite the existing label, the volume is left in an unmounted state. To use this volume, mount it again.

Labeling volumes without mounting

Volumes can be prelabeled without being mounted.

To label a volume without mounting, follow the same procedures as for labeling and mounting in one operation, but clear the Mount After Labeling attribute in the Label window.

Labeling volumes without mounting 195

Backup Storage

Mounting uninventoried volumes

You can mount volumes that are not included in the library inventory, but are valid

(properly labelled) NetWorker volumes.

Procedure

1. In the Administration window, click Devices.

2. Select View

Diagnostic Mode on the toolbar.

3. Manually insert the volume in an empty library slot.

4. Open the Libraries folder in the navigation tree. The Libraries detail table appears.

5. Select the library in the navigation tree in which the volume was manually inserted, or double-click the same library in the Libraries detail table. The

Libraries detail table changes to the double-paned library operations view. The library’s drives are listed in the Devices column, and its slots are listed in the

Slot column.

6. In the Devices column, right-click the library in which the volume was manually inserted, and select Inventory. The Inventory Library window appears.

7. Type the slot number of the volume in both the First and Last field of the Slot

Range.

8. Select Operation Type: either Slow/Verbose (the default) or Fast/Silent.

When Slow/Verbose is selected, the Supply Input option and icon on the

Operations screen of the Monitoring window can be used to confirm the choice to relabel a volume. The device path appears in the Device field.

Entering user input

on page

55 provides details.

9. Click OK.

l l

The Library Operation window displays this message:

The library operation has started.

The Monitoring > Operations screen displays its status.

The NetWorker software then inventories the specified slot.

10. Mount the inventoried volume.

NOTICE

196 EMC NetWorker 9.1 Administration Guide

Backup Storage

Labeling volumes

Several preconfigured label templates are supplied with the NetWorker software. You

cannot delete these preconfigured label templates. Naming label templates

on page 74 provides more information.

When you label a volume, the labeling process: l l l

Writes a label on the volume.

Adds the volume label to the media database.

Prepares tape media to have data written to it.

When you re-label tape, the data on the tape is effectively gone.

During data recovery, the server requests the volume that contains the required data, identifying the required volume by the name with which it was labeled.

Labeling or re-labeling library volumes

Procedure

1. In the Administration window, click Devices.

2. In the left pane, select Libraries.

A list of libraries appears in the right pane.

3. Right-click the library and select Label.

Details for the selected library appear, including divided tables for devices and slots. The Label Library Media dialog box also appears.

4. From the Target Media Pool list, select the pool for the volume.

The pool determines the label template that is used to label the volume.

5. To require manual recycling of the volume, select Allow

Manual Recycle.

With manual recycling, the volume is not automatically marked as recyclable when all save sets expire. You must manually mark the volume as recyclable.

NOTICE

6. To be prompted before the existing label is overwritten, select Prompt to

overwrite label.

7. Click OK.

The Library Operation dialog box appears, stating that the library operation has started.

Labeling volumes 197

Backup Storage

8. To track the status of the label operation, click Monitoring in the

Administration window.

Input.

A confirmation message appears.

b. Click Yes.

Verifying the label when a volume is unloaded

If a SCSI reset is issued during a backup, the volume rewinds and NetWorker may overwrite the volume label.

To detect if the label is overwritten in this circumstance, select the Verify label on

eject checkbox in the Device resource, or set the Verify label on unload setting in the

Troubleshooting devices and autochangers

This section explains how to resolve problems with devices and autochangers.

NOTICE

Do not edit device files and directories, this can result in unpredictable behavior and make it impossible to recover data.

Additional attributes in the Autochanger resource

The Autochanger resource contains attributes that provide a detailed view of the

hidden options that the nsrjb program uses. Displaying diagnostic mode attributes

on page 835 provides information about how to display hidden attributes.

The EMC NetWorker Command Reference Guide and the UNIX man pages provide information about these attributes.

NOTICE

Do not change time related attributes unless advised to do so by a Technical Support representative.

Maintenance commands

NetWorker device driver software provides maintenance commands, such as lusbinfo and lusdebug, that you can use to diagnose problems on tape devices and autochangers.

The EMC NetWorker Command Reference Guide or the UNIX man pages provide information about how to use these commands.

198 EMC NetWorker 9.1 Administration Guide

Backup Storage

Autodetected SCSI jukebox option causes server to stop responding

If you use the jbconfig command to create an autodetected SCSI jukebox and the server stops responding, perform the following steps.

1. Start the jbconfig program

2. Select the option that installs an SJI jukebox.

3. Type the number that corresponds to the type of jukebox you are installing.

4. Continue with jbconfig until this message appears:

Jukebox has been added successfully.

Autochanger inventory problems

This section provides an overview of the situations that can result in an outdated autochanger inventory of volumes and how to update the inventory. When the jukebox inventory becomes outdated, the NetWorker software cannot use the autochanger.

The autochanger inventory can become out of date when: l l

You manually eject the media from the autochanger drive.

You manually remove the media is from the autochanger.

You open the autochanger door.

To update the inventory and enable the NetWorker software to use the autochanger again, perform the following steps.

1. Verify that the volume is correctly installed in the autochanger and that the autochanger door is closed.

2. Log in as root or administrator on the NetWorker server.

3. Reset the autochanger:

nsrjb -Hv

4. Inventory: the autochanger:

nsrjb -Iv

The EMC NetWorker Command Reference Guide or the UNIX man pages provide complete details on the nsrjb command.

Destination component full messages

When you perform a manual operation on an autochanger, for example when you use the buttons on the autochanger to unload the tape drive instead of unloading the tape drive by using NetWorker operations, a message similar to the following may appear:

Destination component full

To resolve the problem, use the nsrjb -H command to reset the autochanger.

Autodetected SCSI jukebox option causes server to stop responding 199

Backup Storage

Tapes do not fill to capacity

The data stored on a tapes may not always fill the tape to capacity. For example, the

NetWorker server can mark a tape with an advertised capacity of 4,000 MB full, after writing only 3,000 MB of data.

To enable NetWorker to use the maximum tape capacity, select the highest density device driver for the device. Additional reasons that the server appears to fill tapes prematurely include: l l l l l

Write errors occur during a backup. With any tape error, the NetWorker server marks the tape as full. To prevent tape write errors, clean the tape drive regularly and use only data-quality tapes. If cleaning the drive does not help, ensure that you perform the following actions: n n n n

Confirm the configuration of the device driver.

Set any necessary switch settings on the tape drive, based on the manufacturer specifications.

Confirm that all cables are secure.

Address other potential SCSI problems.

Space requirements for NetWorker to create file marks. The NetWorker server periodically writes file marks to facilitate rapid recovery of data. These file marks consume varying amounts of tape space, depending on the type of tape drive. The number of file marks the server writes to the tape depends on how many save sets are on the tape. Many small save sets require more file marks than a few larger ones.

Tape capacity differences. Two apparently identical tapes from the same vendor can vary significantly in capacity. This can cause problems when you copy one full tape to another, especially if the destination tape holds less data than the source tape.

Data compression affects the tape capacity. If you use compression on the tape drive, you cannot predict the effect on tape capacity. A compressing drive can provide twice the capacity of a non-compressing drive. Tape capacity can vary depending on the type of backup data. For example, if a non-compressing drive writes 2 GB of data to a specific tape, the compressing drive could write 10 GB, 2

GB, 5 GB, or some other unpredictable amount of data.

Tape length. Verify the tape lengths, for example, a 120-meter DAT tape holds more data than a 90-meter DAT tape.

Tapes get stuck in drive when labeling tapes on Linux Red Hat platform

When you label a tape in a DDS configuration on an RHEL NetWorker server, the tape may become stuck in the drive and display the following error message: unload failure-retrying 30 seconds

To resolve this issue, set the auto_lock setting attribute to “0” (Off) in the /etc/stinit.def file for the following drive types: l l l l

Sony AIT-2 and AIT-3

IBM LTO Gen1

HP LTO Gen1

IBM LTO GEN2

200 EMC NetWorker 9.1 Administration Guide

Backup Storage l l

IBM 3580 drive LTO-1

IBM 3592 J1A l

Quantum DLT 7000

By default the auto_lock setting is set to 1 (On).

Increasing the value of Save Mount Time-out for label operations

A label operation initiated by a backup operation may take more than 30 minutes before it fails when the Auto media management option is enabled and the label operation encounters a corrupted tape.

The NetWorker software keeps a record of the location of the corrupted tape only for the current backup operation, and NetWorker can attempt to use a corrupted tape for the other backup operation, unless an operator removes the volume.

To modify the time it takes the label operation timeout, modify the Save Mount Time-

out attribute for the storage node. Configuring timeouts for storage node remote devices describes how to modify the attribute.

Server cannot access autochanger control port

The control port controls the autochanger loading mechanism. The autochanger hardware installation manual contains information about how to verify that the control port is correctly connected.

If you cannot determine that the control port is working, contact the autochanger vendor for assistance.

Modifying the control port

When a change in the control port of the robotic arm of a library occurs, NetWorker may not be able to perform library operations, such as labeling, mounting, and unmounting, and inventorying. You may see the error no such file or directory when NetWorker tries to perform library operations.

To update the NetWorker server or storage node to use the new control port, perform the following steps.

Procedure

1. Run the inquire command to determine the SCSI device address of the library arm and to confirm that a serial number is reported.

NOTICE

Use the inquire command with caution. The inquire command sends the

SCSI inquiry command to all devices detected on the SCSI bus. If you use inquire during normal operations, unforeseen errors and possible data loss may result.

If inquire reports the serial number of the arm, follow the procedure at

Scanning for libraries and devices on page 143 to scan the library for

devices, then enable the library in NMC: a. In the Administration window, click Devices.

b. Expand the Libraries folder, then right-click the library and select

Enabled/Disable.

Increasing the value of Save Mount Time-out for label operations 201

Backup Storage l

If inquire does not report the serial number or if the scan for devices operation does not detect the control port change, use the nsradmin command to change the control port: a. Log in as root or as Windows administrator on the NetWorker host that manages the control port.

b. At the command prompt, type nsradmin The nsradmin prompt appears.

c. To disable the library, type the following commands:

type: NSR jukebox update enabled: no

d. When nsradmin prompts you to update the resource, type

yes

e. To update the control port, type:

update control port: scsidev@b.t.l

where b.t.l is the bus.target.lun of the library’s robotic arm (as reported by the inquire command).

f. When nsradmin prompts you to update the resource, type

yes

g. To re-enable the library, type:

update enabled: yes

h. When nsradmin prompts you to update the resource, type

yes

i. To verify that the control port was changed and the library is now enabled, type print at the nsradmin prompt.

Changing the sleep times required for TZ89 drive types

When you unload a volume from a TZ89 tape device you may receive an error message similar to the following and NetWorker will repeatedly try to unload the tape: nsrd: media info: unload retry for jukebox `COMPAQTL895' failed

- will retry again.

To resolve this issue, changes the sleep attributes in the Autochanger resource.

1. Shut down NetWorker services.

2. Shut down and restart the autochanger that contains the TZ89 drives.

3. When the autochanger is back online, restart NetWorker services. NetWorker will not try to unload the drive again.

4. Use NMC to edit the following autochanger sleep time attributes, and use the following values: l l

Eject Sleep: 18 secs

Unload Sleep: 40 secs l

Load Sleep: 40 secs

Additional attributes in the Autochanger resource

on page 198 provides information about how to set the sleep attributes.

5. Try to unload the drive again. If the drive fails to unload, repeat this procedure and increase the sleep times.

202 EMC NetWorker 9.1 Administration Guide

Backup Storage

Message displayed when CDI enabled on NDMP or file type device

If you enable the CDI feature for an NDMP tape device or file type device (FTD), a message similar to the following appears: nsrd: media notice: The CDI attribute for device "/dev/rmt/

3cbn" has been changed to "Not used".

To avoid this message, do not enable the CDI attribute for these device types.

Verify firmware for switches and routers

Ensure that the switches or routers firmware that you use on the network was manufactured after August 1995. Most of the switch and router vendors have significantly improved their handling of RPC traffic since August 1995.

Commands issued with nsrjb on a multi-NIC host fail

When you run nsrjb commands to manage a jukebox on a NetWorker server or storage node that has multiple network interface cards (NIC), the commands may fail.

To prevent this failure, add the domain name of each additional NIC to the Aliases attribute in the Client resource for the NetWorker server or storage node.

Editing a

Client resource

on page 422 describes how to edit a Client resource.

SCSI reserve/release with dynamic drive sharing

When the NetWorker software uses Dynamic Drive Sharing (DDS) the operating system tape driver might use the SCSI reserve/release feature in a manner that interferes with the proper operations of the NetWorker software. To resolve this issue, disable the reserve/release feature.

Solaris

The st.conf file contains a setting for each device type in use that enables or disables the SCSI reserve/release feature. The Tape Configuration section of the st man page provides more information. Use the most up-to-date st driver that is available for the version of Solaris.

Edit the st.conf file only if one of the following conditions apply: l l

The NetWorker configuration includes DDS.

Solaris st does not support a tape drive that is configured on a Solaris host.

To determine if the Solaris st tape driver supports a tape drive, perform the following steps:

1. Use the mt command to load a tape in the drive. For example, with the tape device file 0cbn, the type: mt -f /dev/rmt/0cbn status l

If the output of the mt command includes the line SCSI tape drive or appears similar to the following, the st tape driver uses generic settings, which do not support the tape drive:

Message displayed when CDI enabled on NDMP or file type device 203

Backup Storage l mt -f /dev/rmt/4cbn status

Vendor 'IBM ' Product 'ULT3580-TD2 ' tape drive: sense key(0x6)= Unit Attention residual= 0 retries= 0 file no= 0 block no= 0

Tape operations may appear to work in NetWorker but you may run into problems when you try to recover saved data.

If the output of the mt command appears similar to the following, the st tape driver recognizes the drive and uses the correct internal settings to manage the drive: mt -f /dev/rmt/0cbn status

HP Ultrium LTO tape drive: sense key(0x0)= No Additional Sense residual= 0 retries= 0 file no= 0 block no= 0

In this configuration, you must only edit the st.conf file when you use the drive in a DDS configuration.

AIX

To reset the reserve/release setting on an AIX operating system, use the SMIT interface.

1. From the Devices menu, select Tapes.

2. Change the value for the RESERVE/RELEASE support attribute from No to Yes.

HP-UX

To reset the reserve/release setting on an HP-UX 11 operating system, perform the following steps.

1. Change the st_ats_enable kernel variable to a value other than zero.

2. (Optional) Restart the computer to ensure that the operating system implements the change.

Note

The reserve/release is a fixed setting in HP-UX 10.

Recovering save sets from a VTL on a different NetWorker server

The following procedure describes the steps that you need to perform before you can load a tape that was in a VTL managed by one NetWorker server into a different

NetWorker server.

Before you begin

Ensure the destination VTL is the same model, has the same drive names and the same number of drives as the original VTL.

Procedure

1. Confirm the inventory of the VTL in the destination NetWorker storage node

2. Run the inquire command to determine the Control port of the VTL on the destination NetWorker storage node.

3. Run the sjimm command to load the tape into a drive on the destination

NetWorker server.

204 EMC NetWorker 9.1 Administration Guide

Backup Storage

4. Use the mt command to ensure that the tape status is online. For example: mt f device_name status

When the mt command reports that the tape drive is online, you can use the scanner command to scan the save set information into the media database and client file index of the destination NetWorker server.

Recovering save sets from a VTL on a different NetWorker server 205

Backup Storage

206 EMC NetWorker 9.1 Administration Guide

CHAPTER 4 Data Protection Policies

This chapter contains the following topics: l l l l l l l l l l

Overview of data protection policies

................................................................ 208

Designing data protection policies

....................................................................209

Policy notifications

...........................................................................................268

Monitoring policy activity

.................................................................................269

Policy log files

................................................................................................... 271

Starting, stopping, and restarting policies

........................................................ 273

Starting actions in a workflow for an individual client

....................................... 273

Modifying data protection Policy resources

..................................................... 274

Managing policies from the command prompt

..................................................287

Troubleshooting policies

.................................................................................. 295

Data Protection Policies 207

Data Protection Policies

Overview of data protection policies

Data protection policy is a concept that provides you with the ability to design a data protection solution for the environment at the data level instead of at the host level.

With a data protection policy, each client in the environment is a backup object and not simply a host.

Data protection policies enable you to back up and manage data in a variety of environments, as well as to perform system maintenance tasks on the NetWorker server.

A data protection policy solution encompasses the configuration of the following key

NetWorker resources:

Policies

Policies provide you with the ability to develop a service-catalogue approach to the configuration of a NetWorker datazone. Policies enable you to manage all data protection tasks and the data protection lifecycle from a central location.

Policies provide an organizational container for the workflows, actions, and groups that support and define the backup, clone, management, and system maintenance actions that you want to perform.

Workflows

Workflows define the start time for a series of actions, the frequency in which the actions run, the order of actions in a sequence, and the protection group to which the workflow applies.

A workflow can be as simple as a single action that applies to a finite list of Client resources, or a complex chain of actions that apply to a dynamically changing list of resources. In a workflow, some actions can be set to occur sequentially, and others can occur concurrently.

You can create multiple workflows in a single policy. However, each workflow can belong to only one policy. When you add multiple workflows to the same policy, you can logically group data protection activities with similar service level provisions together, to provide easier configuration, access, and task execution.

Protection groups

Protection groups define a set of static or dynamic Client resources or save sets to which a workflow applies. There are also dedicated protection groups for backups in a

VMware environment or for snapshot backups on a NAS device. Review the following information about protection groups: l l l

Create one protection group for each workflow. Each group can be assigned to only one workflow.

You can add the same Client resources and save sets to more than one group at a time.

You can create the group before you create the workflow, or you can create the group after you create the workflow and then assign the group to the workflow later.

Actions

Actions are the key resources in a workflow for a data protection policy and define a specific task, for example, a backup, clone, or snapshot. NetWorker uses a work list to define the task. A work list is composed of one or several work items. Work items include client resources, virtual machines, save sets, or tags. You can chain multiple actions together to occur sequentially or concurrently in a workflow. All chained actions use the same work list.

208 EMC NetWorker 9.1 Administration Guide

Data Protection Policies

When you configure an action, you define the days on which to perform the action, as well as other settings specific to the action. For example, you can specify a destination pool, a retention period, and a target storage node for the backup action, which can differ from the subsequent action that clones the data.

You can create multiple actions for a single workflow. However, each action applies to a single workflow and policy.

The following figure provides a high level overview of the components that make up a data protection policy in a datazone.

Figure 13 Data Protection Policy

Designing data protection policies

Designing and developing effective data protection policies requires thoughtful analysis of the client resources from which to back up data, the actions to perform on the data, and the order and timing of the actions.

Data protection policies can be grouped into six main strategies: l l l l l

Traditional backups—Includes file system backups, NDMP backups, NMDA backups, NMM backups, and Block Based Backups. The EMC NetWorker Network

Data Management Protocol (NDMP) User Guide provides detailed information about how to backup, clone, and recover NDMP data. The NMM and NMDA documentation provides information about how to backup, clone, and recover application data.

NetWorker and NMC Server database backups and maintenance activities—

Performs NetWorker server bootstrap and NMC database backups.

Snapshot backups—Includes snapshot backups of supported EMC storage arrays or appliances. You can clone snapshot data currently with the backup operation, or after the snapshot backup completes. The EMC NetWorker Snapshot Management

Integration Guide describes how to configure data protection policies for EMC storage arrays and appliances with the NetWorker Snapshot Management feature.

NAS device backups—Includes file system snapshots, and NAS snapshots. You can clone data after a snapshot backup job completes or concurrently. The EMC

NetWorker Snapshot Management Integration Guide describes how to configure data protection policies for snapshot backups.

VMware backups—Includes NetWorker VMware Protection with the vProxy appliance (NVP), VMware Backup Appliance (VBA) backups, VBA checkpoint backups for disaster recovery, and virtual machine backups. The EMC NetWorker

VMware Integration Guide describes how to configure data protection policies for

NVP, VBA, VBA checkpoint, and virtual machine backups and clones.

Designing data protection policies 209

Data Protection Policies l

Cloning- You can configure data protection policies that clone backup data by querying the media database for a list of save sets that are based on user defined criteria.

Note

You can also clone traditional, snapshot, bootstrap, and VMware backup data concurrently with the backup operation, or after the backup operation completes.

The Integration Guides provide detailed information about how to clone Snapshot and VMware backup data.

Default data protection policies

NetWorker provides you with preconfigured data protection policy resources that you can use immediately to protect your environment, modify to suit your environment, or use an example to create new resource configurations. To use these policy resources, you must add clients to the appropriate group resource.

Note

NetWorker also includes a preconfigured Server Protection policy to protect the

NetWorker and NMC server databases. The section "Server Protection policy and workflows" provides more information.

Each protection policy provides an example of the EMC best practices that you should follow when you design your data protection solution: l

Separate file system backups from application database backups, to provide ease of access at recovery time.

Stagger the start times for file system backup from the application database backups, to prevent disk contention on the target hosts.

The default data protection policy resources mimic the requirements of a service provider, with different policies that are designed to provide protection based on service level agreements.

Platinum policy

The Platinum policy provides you with an example of a data protection policy for an environment that contains EMC storage arrays or appliances and requires backup data redundancy. The policy contains one workflow with two actions, a snapshot backup action, followed by a clone action.

Figure 14 Platinum policy configuration

Gold policy

The Gold policy provides an example of a data protection policy for an environment that contains virtual machines and requires backup data redundancy. The policy contains two workflows, one to protect Hyper-V hosts and one to protect VMware hosts. Each workflow contains a backup action followed by a clone action.

210 EMC NetWorker 9.1 Administration Guide

Data Protection Policies

Figure 15 Gold policy configuration

Silver policy

The Silver policy provides an example of a data protection policy for an environment that contains non-virtualized machines and requires backup data redundancy. The policy contains two workflows, one to protect hosts file systems and one to protect database applications. Each workflow contains a backup action followed by a clone action.

Figure 16 Silver policy configuration

Bronze policy

The Bronze policy provides an example of a data protection policy for an environment that contains non-virtualized machines. The policy contains two workflows, one to protect hosts file systems and one to protect database applications. Each workflow contains a backup action.

Figure 17 Bronze policy configuration

Road map for configuring a new data protection policy

Procedure

1. Create a policy.

Road map for configuring a new data protection policy 211

Data Protection Policies

When you create a policy, you specify the name and notification settings for the policy.

2. Within the policy, create a workflow for each data type.

For example, create one workflow to protect file system data and one workflow to protect application data. When you create a workflow, you specify the name of the workflow, the time to start the workflow, notification settings for the workflow, and the protection group to which the workflow applies.

3. Create a protection group.

The type of group that you create depends on the types of clients and data that you want to protect. The actions that appear for a group depend on the group type.

4. Create one or more actions for the workflow.

5. To define the backup data that you want to protect, configure Client resources, and then assign the client resources to a protection group.

The following figure illustrates a policy with two different workflows. Workflow 1 performs a probe and then a backup of the Client resources in Client group 1, and then clones the save sets from the backups. Workflow 2 performs a backup of the Client resources in Dynamic client group 1, and then clones the save sets from the backups.

Figure 18 Data protection policy example

NetWorker resource considerations

When you create NetWorker workflow and action resources, consider the following recommendations: l

The parallelism value for the action resource should not exceed 25.

212 EMC NetWorker 9.1 Administration Guide

Data Protection Policies l

The total number of clients in a single workflow should not exceed 100.

Strategies for traditional backups

The primary considerations for a traditional backup strategy are the groups of Client resources, the workflows that define the series of actions that are associated with the backup, and the schedule for the backup.

Creating a policy

Procedure

1. On the Administration window, click Protection.

2. In the expanded left pane, right-click Policies, and then select New.

The Create Policy dialog box appears.

3. On the General tab, in the Name field type a name for the policy.

The maximum number of characters for the policy name is 128.

Note

After you create a policy, the Name attribute is read-only.

4. In the Comment box, type a description for the policy.

5. From the Send Notifications list, select whether to send notifications for the policy: l l

To avoid sending notifications, select Never.

To send notifications with information about each successful and failed workflow and action after all the actions in the policy complete, select On

Completion.

To send a notification with information about each failed workflow and action after all the actions in the policy complete, select On Failure.

6. In the Send notification attribute when you select the On Completion or On

failure option, the Command box appears. Use this box to configure how

NetWorker sends the notifications. You can use the nsrlog action to write the notifications to a log file or configure an email notification.

The default notification action is to log the information to the policy_notifications.log file. The policy_notifications.log file is located in the /nsr/logs directory on Linux and the C:\Program Files

\EMC NetWorker\nsr\logs folder on Windows, by default. You can use the smtpmail application on Windows or the default mailer program on Linux to send email messages.

For example: l

To log notifications to a file named policy_notifications.log, type the following command: l nsrlog -f policy_notifications.log

On Linux, to send a notification email, type the following command: l mail -s subject recipient

For NetWorker Virtual Edition (NVE), to send a notification email, type the following command:

Strategies for traditional backups 213

Data Protection Policies l

/usr/sbin/sendmail -v recipient_email "subject_text"

On Windows, to send a notification email, type the following command: smtpmail -s subject -h mailserver recipient1@mailserver

recipient2@mailserver...

where: n n n

-s subject

—Includes a standard email header with the message and specifies the subject text for that header. Without this option, the smtpmail program assumes that the message contains a correctly formatted email header and nothing is added.

-h mailserver

—Specifies the hostname of the mail server to use to relay the SMTP email message.

recipient1@mailserver—Is the email address of the recipient of the notification. Multiple email recipients are separated by a space.

7. To specify the Restricted Data Zone (RDZ) for the policy, select the Restricted

Data Zones tab, and then select the RDZ from the list.

8. Click OK.

After you finish

Create the workflows and actions for the policy.

Creating a workflow in a new policy

A policy must contain one or more workflows.

Procedure

1. In the left pane of the Protection window, expand Policies, and then select the policy that you created.

2. In the right pane of the Protection window, select Create a new workflow.

3. In the Name field, type the name of the workflow.

The maximum number of allowed characters for the Name field is 64.

4. In the Comment box, type a description for the workflow. The maximum number of allowed characters for the Comment field is 128.

5. From the Send Notifications list, select how to send notifications for the workflow: l l

To use the notification configuration that is defined in the policy resource to determine when to send the notification, select Set at policy level.

To send notifications with information about each successful and failed workflow and action, after all the actions in the workflow complete, select

On Completion.

To send notifications with information about each failed workflow and action, after all the actions in the workflow complete, select On Failure.

6. In the Send notification attribute when you select the On Completion or On

failure option, the Command box appears. Use this box to configure how

NetWorker sends the notifications. You can use the nsrlog action to write the notifications to a log file or configure an email notification.

214 EMC NetWorker 9.1 Administration Guide

Data Protection Policies

\EMC NetWorker\nsr\logs folder on Windows, by default. You can use the smtpmail application on Windows or the default mailer program on Linux to send email messages.

For example: l

To log notifications to a file named policy_notifications.log, type the following command: l nsrlog -f policy_notifications.log

On Linux, to send a notification email, type the following command: l mail -s subject recipient

For NetWorker Virtual Edition (NVE), to send a notification email, type the following command: l

/usr/sbin/sendmail -v recipient_email "subject_text"

On Windows, type the following command: smtpmail -s subject -h mailserver recipient1@mailserver

recipient2@mailserver...

where: n n n

-s subject

-h mailserver

—Specifies the hostname of the mail server to use to relay the SMTP email message.

recipient1@mailserver—Is the email address of the recipient of the notification. Multiple email recipients are separated by a space.

7. In the Running group box, define when and how often the workflow runs.

a. To ensure that the actions contained in the workflow run when the policy or workflow is started, in the Enabled box, leave the option selected. To prevent the actions in the workflow from running when the policy or workflow that contains the action is started, clear this option.

b. To ensure that the workflow starts at the time that is specified in the Start

time attribute, on the days that are defined in the action resource, in the

AutoStart Enabled box, leave the option selected. To prevent the workflow from running at the time that is specified in the Start time attribute, clear this option.

c. To define the time to start the actions in the workflow, in the Start Time attribute, use the spin boxes.

The default value is 9:00 P.M.

d. To define how frequently to repeat the actions that are defined in the workflow over a 24 hour period, In the Interval attribute, use the spin boxes.

The default value is 24 hours, or once a day. When you select a value that is less than 24 hours, the Interval End attribute appears. To define the last time to start a workflow in a defined interval period, use the spin boxes.

Strategies for traditional backups 215

Data Protection Policies e. To define the duration of time in which NetWorker can manually or automatically restart a failed or canceled workflow, in the Restart Window attribute, use the spin boxes.

If the restart window has elapsed, NetWorker considers the restart as a new run of the workflow. NetWorker calculates the restart window from the start of the last incomplete workflow. The default value is 24 hours.

For example, when you set the Start Time to 7:00 PM, the Interval to 1 hour, and the Interval end time to 11:00 P.M., then the workflow automatically starts every hour beginning at 7:00 P.M. and the last start time is 11:00 PM.

8. To create the workflow, click OK.

After you finish

Create the actions that will occur in the workflow, and then assign a group to the workflow. If a workflow does not contain a group, a policy does not perform any actions.

Protection groups for traditional backups

Create protection groups for traditional backups, which identifies the Client resources to back up.

You can create two types of protection groups for a traditional backup: l l

Basic client group—Defines a static list of Client resources to back up.

Dynamic client group—Specifies a dynamic list of Client resources to back up. A dynamic client group automatically generates a list of Client resources that use client tag which matches the client tag that is specified for the group.

Create multiple groups to perform different types of backups for different Client resources, or to perform backups on different schedules. For example: l l l

Create one group for backups of clients in the Accounting department, and another group for backups of clients in the Marketing department.

Create one group for file system backups and one group for backups of Microsoft

Exchange data with the NetWorker Module for Microsoft.

Create one group for a workflow with backups actions that start at 11 p.m., and another group for a workflow with backup actions that start at 2 a.m.

216

Note

A Client resource can belong to more than one group.

Creating a client group

Basic client groups define a static list of Client resources for a traditional backup, check connectivity, or probe action.

Before you begin

Create the Client resources for the data to include in a protection group.

Procedure

1. In the Administration window, click Protection.

2. In the expanded left pane, right-click Groups, and then select New.

The Create Group dialog box appears, starting with the General tab.

3. In the Name attribute, type a name for the group.

EMC NetWorker 9.1 Administration Guide

Data Protection Policies

The maximum number of characters for the group name is 64.

Note

After you create a group, the Name attribute is read-only.

4. From the Group Type list, leave the default selection of Clients.

5. In the Comment field, type a description of the group.

6. Select the workflow in which to assign the group from the Policy-Workflow list.

Note

You can also assign the group to a workflow when you create or edit a workflow.

7. (Optional) On the Restricted Datazones tab, to specify the Restricted

Datazone (RDZ) for the group, select the RDZ from the list.

8. Click OK.

After you finish

Create Client resources. The Client Configuration wizard and General tab on the

Client Properties dialog box properties page provide you with the ability to assign clients to a protection group.

Creating a dynamic client group

Dynamic client groups prevent you from having to edit group settings when you add

Client resources to the NetWorker datazone. You can configure a dynamic group to include all the clients on the NetWorker server or you can configure the dynamic group to perform a query that generates a list of clients that is based on a matching tag value. A tag is a string attribute that you define in a Client resource. When an action starts in a workflow that is a member of a tagged dynamic protection group, the policy engine dynamically generates a list of Client resources that match the tag attribute value.

Use dynamic client groups to specify a dynamic list of Client resources for a traditional backup, probe, check connectivity, or server backup action.

Procedure

1. In the Administration window, click Protection.

2. In the expanded left pane, right-click Groups, and then select New.

The Create Group dialog box appears, starting with the General tab.

3. In the Name attribute, type a name for the group.

The maximum number of characters for the group name is 64.

Note

After you create a group, the Name attribute is read-only.

4. From the Group Type list, select Dynamic Clients.

5. In the Comment field, type a description of the group.

6. Select the workflow in which to assign the group from the Policy-Workflow list.

Strategies for traditional backups 217

Data Protection Policies

Note

You can also assign the group to a workflow when you create or edit a workflow.

7. (Optional) On the Restricted Datazones tab, to specify the Restricted

Datazone (RDZ) for the group, select the RDZ from the list.

8. Click OK.

After you finish

Create Client resources. The Client Configuration wizard and General tab on the

Client Properties dialog box properties page provide you with the ability to assign clients to a protection group and define one or more tags.

Supported actions in traditional backup workflows

Traditional backup workflows can optionally include a probe or check connectivity action before the backup, and a clone action either concurrently with or after the backup.

Probe

A probe action runs a user-defined script on a NetWorker client before the start of a backup. A user-defined script is any program that passes a return code. If the return code is 0 (zero), then a client backup is required. If the return code is 1, then a client backup is not required.

Only a backup action can follow a probe action.

Check connectivity

A check connectivity action tests connectivity between clients and the NetWorker server before a probe or backup action occurs. If the connectivity test fails, then the backup does not occur on the client.

Traditional backup

A traditional backup is a scheduled backup of the save sets defined for the Client resources in the assigned group. You must specify the destination storage node, destination pool, the schedule (period and activity), and the retention period for the backup.

Clone

A clone action creates a copy of one or more save sets. Cloning allows for secure offsite storage, transfer of data from one location to another, and verification of backups.

You can configure a clone action to occur after a backup in a single workflow, or simultaneously with a backup action in a single workflow. You can also use save set and query groups to define a specific list of save sets to clone, in a separate workflow.

Actions sequences in traditional backup workflows

Workflows enable you to chain together multiple actions and run them sequentially or concurrently.

A workflow for a traditional backup can optionally include a probe or check connectivity action before the backup, and a clone action either concurrently with or after the backup.

The following sections provide details on supported actions that can follow the lead action and other actions in a workflow.

218 EMC NetWorker 9.1 Administration Guide

Data Protection Policies

All possible workflow actions for a traditional backup

The following figure illustrates the possible workflow actions that are associated with a traditional backup.

Figure 19 All possible workflow actions for a traditional backup

Workflow path from a traditional backup action

The only action that can follow a traditional backup is a clone action.

Figure 20 Workflow path from a traditional backup action

Creating a check connectivity action

A check connectivity action tests connectivity between clients and the NetWorker server, usually before another action such as a backup occurs.

Before you begin

Create the policy and workflow that contain the action. The check connectivity action should be the first action in the workflow.

Procedure

1. In the expanded left pane, select the workflow, and then perform one of the following tasks in the right pane to start the Policy Action wizard: l l

If this is the first action in the workflow, select Create a new action.

If the workflow has other actions, right-click an empty area of the Actions pane, and then select New.

The Specify the Action Information page appears.

2. In the Name field, type the name of the action.

The maximum number of characters for the action name is 64.

3. In the Comment field, type a description for the action.

4. To ensure that the action runs when the policy or workflow that contains the action is started, in the Enabled box, select the option. To prevent the action from running when the policy or workflow that contains the action is started, clear this option.

Note

When you clear the Enabled option, any action that occurs after a disabled action will not start, even if the succeeding options are enabled.

5. From the Action Type list, select Check Connectivity.

Strategies for traditional backups 219

Data Protection Policies

6. When you create the action as part of the workflow configuration, the workflow appears automatically in the Workflow box and the box is grayed out.

7. Specify the order of the action in relation to other actions in the workflow: l

If the action is part of a sequence of actions in a workflow path, select the action that should precede this action from the Previous box.

If the action should run concurrently with an action, select the concurrent action from the Previous box, and then select the Concurrent checkbox.

8. Select whether to use a weekly or monthly schedule for the action: l

To specify a schedule for each day of the week, select Weekly by day.

To specify a schedule for each day of the month, select Monthly by day.

9. Click the icon on each day to specify whether to check connectivity with the client.

The following table provides details on the icons.

Table 37 Schedule icons

Icon Label

Execute

Skip

Description

Check connectivity on this day.

Do not check connectivity on this day.

To check connectivity every day, select Execute from the list, and then click

Make All.

10. Click Next.

The Specify the Connectivity Options page appears.

11. Select the success criteria for the action: l

To specify that the connectivity check is successful only if successful connectivity is achieved with all clients in the assigned group, select the

Succeed only after all clients succeed checkbox.

To specify that the connectivity check is successful if connectivity is achieved with one or more clients in the assigned group, clear the checkbox.

12. Click Next.

The Specify the Advanced Options page appears.

13. (Optional) Configure advanced options and schedule overrides.

Note

Although the Retries, Retry Delay, Inactivity Timeout, or the Send

Notification options appear, the Check Connectivity action does not support these options and ignores the values.

14. In the Parallelism field, specify the maximum number of concurrent operations for the action.

220 EMC NetWorker 9.1 Administration Guide

Data Protection Policies

Note

The Parallelism value should not exceed 25.

15. From the Failure Impact list, specify what to do when a job fails: l l

To continue the workflow when there are job failures, select Continue.

To abort the current action if there is a failure with one of the jobs, but continue with subsequent actions in the workflow, select Abort action.

Note

The Abort action option only applies to probe actions, and the backup actions for the Traditional and Snapshot action types.

To abort the entire workflow if there is a failure with one of the jobs in the action, select Abort workflow.

Note

If any of the actions fail in the workflow, the workflow status does not appear as interrupted or cancelled. NetWorker reports the workflow status as failed.

16. From the Soft Limit list, specify the amount of time after the action starts to stop the initiation of new activities. The default value of 0 (zero) indicates no limit.

17. From the Hard Limit list, specify the amount of time after the action starts to begin terminating activities. The default value of 0 (zero) indicates no limit.

18. Optional, in Start Time specify the time to start the action.

Use the spin boxes to set the hour and minute values, and select one of the following options from the drop-down list: l l l

Disabled—Do not enforce an action start time. The action will start at the time defined by the workflow.

Absolute—Start the action at the time specified by the values in the spin boxes.

Relative—Start the action after the period of time defined in the spin boxes has elapsed after the start of the workflow.

19. Optional, configure overrides for the task that is scheduled on a specific day.

To change the month on which to schedule the override, use the navigation buttons and the month list box. To change the year, use the spin boxes. You can set an override in the following ways: l l

Select the day in the calendar, which changes the action task for the specific day.

Use the action task list to select the task, then perform one of the following steps: n n

To define an override that occurs on a specific day of the week, every week, select Specified day, then use the drop downs. Click Add Rules

based override.

To perform the action task on the last day of the calendar month, select

Last day of the month. Click Add Rules based override.

Strategies for traditional backups 221

Data Protection Policies

222 n

In the Override field, type an override.

Note

To remove an override, delete the entry from the Override field.

20. Click Next.

The Action Configuration Summary page appears.

21. Review the settings that you specified for the action, and then click Configure.

After you finish

(Optional) Create one of the following actions to automatically occur after the check connectivity action: l l

Probe

Traditional backup l

Note

This option is not available for NAS snapshot backups.

Snapshot backup

Creating a probe action

Before you begin l l l

Create the Probe resource script on the clients that use the probe. Create a client

Probe resource on the NetWorker server, and then associate the client Probe resource with the Client resource on the NetWorker server.

Create the policy and workflow that contain the action.

(Optional) Create a check connectivity action to precede the probe action in the workflow. A check connectivity action is the only supported action that can precede a probe action in a workflow.

Procedure

1. In the expanded left pane, select the workflow, and then perform one of the following tasks in the right pane to start the Policy Action wizard: l l

If this is the first action in the workflow, select Create a new action.

If the workflow has other actions, right-click an empty area of the Actions pane, and then select New.

The Specify the Action Information page appears.

2. In the Name field, type the name of the action.

The maximum number of characters for the action name is 64.

3. In the Comment field, type a description for the action.

EMC NetWorker 9.1 Administration Guide

Data Protection Policies

Note

When you clear the Enabled option, any action that occurs after a disabled action will not start, even if the succeeding options are enabled.

5. From the Action Type list, select Probe.

6. When you create the action as part of the workflow configuration, the workflow appears automatically in the Workflow box and the box is grayed out.

7. Specify the order of the action in relation to other actions in the workflow: l l

If the action is part of a sequence of actions in a workflow path, select the action that should precede this action from the Previous box.

If the action should run concurrently with an action, select the concurrent action from the Previous box, and then select the Concurrent checkbox.

8. Select whether to use a weekly or monthly schedule for the action: l l

To specify a schedule for each day of the week, select Weekly by day.

To specify a schedule for each day of the month, select Monthly by day.

9. Click the icon on each day to specify whether to probe the client.

The following table provides details on the icons.

Table 38 Schedule icons

Icon Label

Execute

Skip

Description

Perform the probe on this day.

Do not perform a probe on this day.

To perform a probe every day, select Execute from the list, and then click

Make All.

10. Click Next.

The Specify the Probe Options page appears.

11. Choose whether to start the subsequent backup action only after all probes succeed by selecting or clearing the Start backup only after all probes

succeed checkbox: l l

To start the backup only if all the probes associated with Client resources in the assigned group succeed, select the checkbox.

To start the backup if any one of the probes are associated with a Client resource in the assigned group succeed, clear the checkbox.

12. Click Next.

The Specify the Advanced Options page appears.

13. In the Retries box, specify the number of times that NetWorker should retry a failed probe or backup action, before NetWorker considers the action as failed.

When the Retries value is 0, NetWorker will not retry a failed backup or probe action.

Strategies for traditional backups 223

Data Protection Policies

Note

The Retries option only applies to probe actions, and the backup actions for the

Traditional and Snapshot action types. When you specify a value for this option in other actions, NetWorker ignores the values.

14. In the Retry Delay field, specify a delay in seconds to wait before retrying a failed backup or probe action. When the Retry Delay value is 0, NetWorker retries the failed backup or probe action immediately.

Note

The Retry Delay option only applies to probe actions, and the backup actions for the Traditional and Snapshot action types. When you specify a value for this option in other actions, NetWorker ignores the values.

15. In the Inactivity Timeout field, specify the maximum number of minutes that a job run by an action is allowed to fail to communicate back to the server.

If the job fails to respond within the timeout value, the server considers the job a failure. If a job fails, NetWorker retries the job immediately. This ensures that no time is lost due to failures.

Increase the timeout value if a backup consistently aborts due to inactivity.

Inactivity timeouts may occur for backups of large save sets, backups of save sets with large sparse files, and incremental backups of many small static files.

Note

The Inactivity Timeout option only applies to probe actions, and the backup actions for the Traditional and Snapshot action types. When you specify a value for this option in other actions, NetWorker ignores the values.

16. In the Parallelism field, specify the maximum number of concurrent operations for the action.

Note

The Parallelism value should not exceed 25.

17. From the Failure Impact list, specify what to do when a job fails: l l

To continue the workflow when there are job failures, select Continue.

To abort the current action if there is a failure with one of the jobs, but continue with subsequent actions in the workflow, select Abort action.

Note

The Abort action option only applies to probe actions, and the backup actions for the Traditional and Snapshot action types.

To abort the entire workflow if there is a failure with one of the jobs in the action, select Abort workflow.

224 EMC NetWorker 9.1 Administration Guide

Data Protection Policies

Note

If any of the actions fail in the workflow, the workflow status does not appear as interrupted or cancelled. NetWorker reports the workflow status as failed.

18. Leave the default selections for the Notification group box. NetWorker does not support notifications for probe actions and ignores the values that are defined in the attributes.

19. From the Soft Limit list, specify the amount of time after the action starts to stop the initiation of new activities. The default value of 0 (zero) indicates no limit.

20. From the Hard Limit list, specify the amount of time after the action starts to begin terminating activities. The default value of 0 (zero) indicates no limit.

21. Optional, in Start Time specify the time to start the action.

Use the spin boxes to set the hour and minute values, and select one of the following options from the drop-down list: l l l

Disabled—Do not enforce an action start time. The action will start at the time defined by the workflow.

Absolute—Start the action at the time specified by the values in the spin boxes.

Relative—Start the action after the period of time defined in the spin boxes has elapsed after the start of the workflow.

22. Optional, configure overrides for the task that is scheduled on a specific day.

To change the month on which to schedule the override, use the navigation buttons and the month list box. To change the year, use the spin boxes. You can set an override in the following ways: l l

Select the day in the calendar, which changes the action task for the specific day.

Use the action task list to select the task, then perform one of the following steps: n n n

To define an override that occurs on a specific day of the week, every week, select Specified day, then use the drop downs. Click Add Rules

based override.

To perform the action task on the last day of the calendar month, select

Last day of the month. Click Add Rules based override.

In the Override field, type an override.

Note

To remove an override, delete the entry from the Override field.

23. Click Next.

The Action Configuration Summary page appears.

24. Review the settings that you specified for the action, and then click Configure.

Strategies for traditional backups 225

Data Protection Policies

Creating a traditional backup action

A traditional backup is a scheduled backup of the save sets defined for the Client resources in the assigned group for the workflow.

Before you begin l l

Create the policy and workflow that contain the action.

(Optional) Create actions to precede the backup action in the workflow.

Supported actions that can precede a backup include: n n

Probe

Check connectivity

Procedure

1. In the expanded left pane, select the workflow, and then perform one of the following tasks in the right pane to start the Policy Action wizard: l l

If this is the first action in the workflow, select Create a new action.

If the workflow has other actions, right-click an empty area of the Actions pane, and then select New.

The Specify the Action Information page appears.

2. In the Name field, type the name of the action.

The maximum number of characters for the action name is 64.

3. In the Comment field, type a description for the action.

Note

When you clear the Enabled option, any action that occurs after a disabled action will not start, even if the succeeding options are enabled.

5. From the Action Type list, select Backup.

6. From the secondary action list, select the backup type, for example,

Traditional.

7. Optional, from the Force Backup Level list select a backup level.

For workflows that have more than one scheduled backup within a 24-hour period, use the Force Backup Level attribute to allow more than one backup to occur at two different backup levels in a 24-hour period. When you select a backup level in the Force Backup Level attribute, the first backup is performed at the scheduled backup level. Each subsequent occurrence of the backup action in the next 24 hours occurs at the level defined in the Force Backup

Level attribute. For example, if the level defined by the schedule is Full and the

Force Backup Level attribute is set to Incr, the first backup started by the action occurs at a level full and subsequent backups, within 24 hours of the start of the full backup are incremental. By default this option is cleared, which means that if the action runs multiple backup operations in a 24 period, all the backups occur at the scheduled backup level.

226 EMC NetWorker 9.1 Administration Guide

Data Protection Policies

8. When you create the action as part of the workflow configuration, the workflow appears automatically in the Workflow box and the box is grayed out.

9. Specify the order of the action in relation to other actions in the workflow: l l

If the action is part of a sequence of actions in a workflow path, select the action that should precede this action from the Previous box.

If the action should run concurrently with an action, select the concurrent action from the Previous box, and then select the Concurrent checkbox.

10. Select whether to use a weekly or monthly schedule for the action: l l

To specify a schedule for each day of the week, select Weekly by day.

To specify a schedule for each day of the month, select Monthly by day.

11. Click the icon on each day to specify the backup level to perform.

The following table provides details about the backup level that each icon represents.

Table 39 Schedule icons

Icon Label

Full

Incr

Cumulative Incr

Logs Only

Incremental Synthetic

Full

Skip

Description

Perform a full backup on this day. Full backups include all files, regardless of whether the files changed.

Perform an incremental backup on this day.

Incremental backups include files that have changed since the last backup of any type

(full or incremental).

Perform a cumulative incremental backup.

Cumulative incremental backups include files that have changed since the last full backup.

Perform a backup of only database transaction logs.

Perform an incremental synthetic backup on this day.

An incremental synthetic full backup includes all data that changed since the last full backup and subsequent incremental backups to create a synthetic full backup.

Do not perform a backup on this day.

To perform the same type of backup on each day, select the backup type from the list and click Make All.

Strategies for traditional backups 227

Data Protection Policies

228

12. Click Next.

The Specify the Backup Options page appears.

13. From the Destination Storage Node box, select the storage node with the devices on which to store the backup data.

14. From the Destination Pool box, select the media pool in which to store the backup data.

15. From the Retention boxes, specify the amount of time to retain the backup data.

After the retention period expires, the save set is removed from the client file index and marked as recyclable in the media database during an expiration server maintenance task.

16. From the Client Override Behavior box, specify how NetWorker uses certain client configuration attributes that perform the same function as attributes in the Action resource.

l l l

Client Can Override—The values in the Client resource for Schedule, Pool,

Retention policy, and the Storage Node attributes take precedence over the values that are defined in the equivalent Action resource attributes.

Client Can Not Override —The values in the Action resource for the

Schedule, Destination Pool, Destination Storage Node, and the

Retention attributes take precedence over the values that are defined in the equivalent Client resource attributes.

Legacy Backup Rules—This value only appears in actions that are created by the migration process. The updating process sets the Client Override

Behavior for the migrated backup actions to Legacy Backup Rules.

17. Click Next.

The Specify the Advanced Options page appears.

18. In the Retries box, specify the number of times that NetWorker should retry a failed probe or backup action, before NetWorker considers the action as failed.

When the Retries value is 0, NetWorker will not retry a failed backup or probe action.

Note

The Retries option only applies to probe actions, and the backup actions for the

Traditional and Snapshot action types. When you specify a value for this option in other actions, NetWorker ignores the values.

19. In the Retry Delay field, specify a delay in seconds to wait before retrying a failed backup or probe action. When the Retry Delay value is 0, NetWorker retries the failed backup or probe action immediately.

Note

20. In the Inactivity Timeout field, specify the maximum number of minutes that a job run by an action is allowed to fail to communicate back to the server.

EMC NetWorker 9.1 Administration Guide

Data Protection Policies

If the job fails to respond within the timeout value, the server considers the job a failure. If a job fails, NetWorker retries the job immediately. This ensures that no time is lost due to failures.

Increase the timeout value if a backup consistently aborts due to inactivity.

Inactivity timeouts may occur for backups of large save sets, backups of save sets with large sparse files, and incremental backups of many small static files.

Note

21. In the Parallelism field, specify the maximum number of concurrent operations for the action.

Note

The Parallelism value should not exceed 25.

22. From the Failure Impact list, specify what to do when a job fails: l l

To continue the workflow when there are job failures, select Continue.

To abort the current action if there is a failure with one of the jobs, but continue with subsequent actions in the workflow, select Abort action.

Note

The Abort action option only applies to probe actions, and the backup actions for the Traditional and Snapshot action types.

To abort the entire workflow if there is a failure with one of the jobs in the action, select Abort workflow.

Note

If any of the actions fail in the workflow, the workflow status does not appear as interrupted or cancelled. NetWorker reports the workflow status as failed.

23. From the Soft Limit list, specify the amount of time after the action starts to stop the initiation of new activities. The default value of 0 (zero) indicates no limit.

24. From the Hard Limit list, specify the amount of time after the action starts to begin terminating activities. The default value of 0 (zero) indicates no limit.

25. Optional, in Start Time specify the time to start the action.

Use the spin boxes to set the hour and minute values, and select one of the following options from the drop-down list: l l l

Disabled—Do not enforce an action start time. The action will start at the time defined by the workflow.

Absolute—Start the action at the time specified by the values in the spin boxes.

Relative—Start the action after the period of time defined in the spin boxes has elapsed after the start of the workflow.

Strategies for traditional backups 229

Data Protection Policies

26. Optional, configure overrides for the task that is scheduled on a specific day.

To change the month on which to schedule the override, use the navigation buttons and the month list box. To change the year, use the spin boxes. You can set an override in the following ways: l l

Select the day in the calendar, which changes the action task for the specific day.

Use the action task list to select the task, then perform one of the following steps: n n n

To define an override that occurs on a specific day of the week, every week, select Specified day, then use the drop downs. Click Add Rules

based override.

To perform the action task on the last day of the calendar month, select

Last day of the month. Click Add Rules based override.

In the Override field, type an override.

Note

To remove an override, delete the entry from the Override field.

27. From the Send Notifications list box, select whether to send notifications for the action: l l l

Select Set at policy level to use the notification configuration that is defined in the Policy resource to send the notification.

Select On Completion to send a notification on completion of the action.

Select On Failure to send a notification only if the action fails to complete.

28. In the Send notification attribute when you select the On Completion or On

failure option, the Command box appears. Use this box to configure how

NetWorker sends the notifications. You can use the nsrlog action to write the notifications to a log file or configure an email notification.

\EMC NetWorker\nsr\logs folder on Windows, by default. You can use the smtpmail application on Windows or the default mailer program on Linux to send email messages.

For example: l

To log notifications to a file named policy_notifications.log, type the following command: l nsrlog -f policy_notifications.log

On Linux, to send a notification email, type the following command: l mail -s subject recipient

For NetWorker Virtual Edition (NVE), to send a notification email, type the following command: l

/usr/sbin/sendmail -v recipient_email "subject_text"

On Window, to send a notification email, type the following command:

230 EMC NetWorker 9.1 Administration Guide

Data Protection Policies smtpmail -s subject -h mailserver recipient1@mailserver

recipient2@mailserver...

where: n n n

-s subject

-h mailserver

—Specifies the hostname of the mail server to use to relay the SMTP email message.

recipient1@mailserver—Is the email address of the recipient of the notification. Multiple email recipients are separated by a space.

29. Click Next.

The Action Configuration Summary page appears.

30. Review the settings that you specified for the action, and then click Configure.

After you finish

(Optional) Create a clone action to automatically clone the save sets after the backup.

A clone action is the only supported action after a backup action in a workflow.

Creating a clone action

A clone action creates a copy of one or more save sets. Cloning allows for secure offsite storage, the transfer of data from one location to another, and the verification of backups.

Procedure

1. In the expanded left pane, select the workflow, and then perform one of the following tasks in the right pane to start the Policy Action wizard: l

If this is the first action in the workflow, select Create a new action.

If the workflow has other actions, right-click an empty area of the Actions pane, and then select New.

The Specify the Action Information page appears.

2. In the Name field, type the name of the action.

The maximum number of characters for the action name is 64.

3. In the Comment field, type a description for the action.

Note

When you clear the Enabled option, any action that occurs after a disabled action will not start, even if the succeeding options are enabled.

5. From the Action Type list, select Clone.

6. When you create the action as part of the workflow configuration, the workflow appears automatically in the Workflow box and the box is grayed out.

Strategies for traditional backups 231

Data Protection Policies

7. Specify the order of the action in relation to other actions in the workflow: l l

If the action is part of a sequence of actions in a workflow path, select the action that should precede this action from the Previous box.

If the action should run concurrently with an action, select the concurrent action from the Previous box, and then select the Concurrent checkbox.

8. Select whether to use a weekly or monthly schedule for the action: l l

To specify a schedule for each day of the week, select Weekly by day.

To specify a schedule for each day of the month, select Monthly by day.

9. Click the icon on each day to specify whether to perform cloning.

The following table provides details on the icons.

Table 40 Schedule icons

Icon Label

Execute

Description

Perform cloning on this day.

232

Skip Do not perform cloning on this day.

To perform cloning every day, select Execute from the list and click Make All.

10. Click Next.

The Specify the Clone Options page appears.

11. In the Data Movement group box, define the volumes and devices to which

NetWorker sends the clone data.

a. From the Destination Storage Node list, select the storage node with the devices on which to store the cloned save sets.

b. In the Delete source save sets after clone completes, select the option to instruct NetWorker to move the data from the source volume to the destination volume after clone operation completes. This is equivalent to staging the save sets.

c. From the Destination Pool list, select the target media pool for the cloned save sets.

d. From the Retention list, specify the amount of time to retain the cloned save sets.

After the retention period expires, the save sets are marked as recyclable during an expiration server maintenance task.

12. In the Filters group box, define the criteria that NetWorker uses to create the list of eligible save sets to clone. The eligible save sets must match the requirements that are defined in each filter. NetWorker provides the following filter options: a. Time filter—Use the Time section to define the time range in which

NetWorker should inspect, when searching for eligible save sets to clone in the media database. Use the spin boxes to specify the start of the time range and the end of the time range. The Time filter list includes three options, which define how NetWorker determines save set eligibility, based on the time criteria:

EMC NetWorker 9.1 Administration Guide

Data Protection Policies l l

Do Not Filter—NetWorker inspects the save sets in the media database to create a clone save set list that meets the filter criteria.

Accept—The clone save set list includes save sets whose save time is within the time range that is specified by the spin boxes and meet all the other defined filter criteria.

Reject—The clone save set list does not include save sets whose save time is within the time range that is specified by the spin boxes and meet all the other defined filter criteria.

b. Save Set filter—Use the Save Set section to instruct NetWorker to include or exclude ProtectPoint and Snapshot save sets, when searching for eligible save sets to clone in the media database. The Save Set filter list includes three options, which define how NetWorker determines save set eligibility, based on the save set criteria: l l l

Do Not Filter—NetWorker inspects the save sets in the media database to create a clone save set list that meets the filter criteria.

Accept—The clone save set list includes eligible ProtectPoint or

Snapshot save sets, when you also enable the ProtectPoint or Snapshot checkboxes.

Reject—The clone save set list does not include eligible ProtectPoint and Snapshot save sets when you also enable the ProtectPoint and

Snapshot checkboxes.

c. Clients filter—Use the Client section to define a list of clients to include or exclude, when NetWorker searches for eligible save sets to clone in the media database. The Client list includes three options, which define how

NetWorker determines save set eligibility, based on the client criteria: l l

Do Not Filter—NetWorker inspects save sets that are associated with the clients in the media database, to create a clone save set list that meets the filter criteria.

Accept—The clone save set list includes eligible save sets for the selected clients.

Reject—The clone save set list does not include eligible save sets for the selected clients.

d. Levels filter—Use the Levels section to define a list of backup levels to include or exclude, when NetWorker searches for eligible save sets to clone in the media database. The Levels filter list includes three options, which define how NetWorker determines save set eligibility, based on the level criteria: l l l

Do Not Filter—NetWorker inspects save sets regardless of level in the media database, to create a clone save set list that meets all the filter criteria.

Accept—The clone save set list includes eligible save sets with the selected backup levels.

Reject—The clone save set list does not include eligible save sets with the selected backup levels.

13. Click Next.

The Specify the Advanced Options page appears.

14. Configure advanced options, including notifications and schedule overrides.

Strategies for traditional backups 233

Data Protection Policies

Note

Although the Retries, Retry Delay, or the Inactivity Timeout options appear, the clone action does not support these options and ignores the values.

15. In the Parallelism field, specify the maximum number of concurrent operations for the action.

Note

The Parallelism value should not exceed 25.

16. From the Failure Impact list, specify what to do when a job fails: l l

To continue the workflow when there are job failures, select Continue.

To abort the current action if there is a failure with one of the jobs, but continue with subsequent actions in the workflow, select Abort action.

Note

The Abort action option only applies to probe actions, and the backup actions for the Traditional and Snapshot action types.

To abort the entire workflow if there is a failure with one of the jobs in the action, select Abort workflow.

234

Note

If any of the actions fail in the workflow, the workflow status does not appear as interrupted or cancelled. NetWorker reports the workflow status as failed.

17. From the Send Notifications list box, select whether to send notifications for the action: l l l

Select Set at policy level to use the notification configuration that is defined in the Policy resource to send the notification.

Select On Completion to send a notification on completion of the action.

Select On Failure to send a notification only if the action fails to complete.

18. In the Send notification attribute when you select the On Completion or On

failure option, the Command box appears. Use this box to configure how

NetWorker sends the notifications. You can use the nsrlog action to write the notifications to a log file or configure an email notification.

\EMC NetWorker\nsr\logs folder on Windows, by default. You can use the smtpmail application on Windows or the default mailer program on Linux to send email messages.

For example: l

To log notifications to a file named policy_notifications.log, type the following command: nsrlog -f policy_notifications.log

EMC NetWorker 9.1 Administration Guide

Data Protection Policies l

On Linux, to send a notification email, type the following command: l mail -s subject recipient

For NetWorker Virtual Edition (NVE), to send a notification email, type the following command: l

/usr/sbin/sendmail -v recipient_email "subject_text"

On Window, to send a notification email, type the following command: smtpmail -s subject -h mailserver recipient1@mailserver