Hitachi Unified Compute Platform 2000 for Microsoft ® Private Cloud Reference Architecture Guide

Hitachi Unified Compute Platform 2000 for Microsoft ® Private Cloud Reference Architecture Guide
Hitachi Unified Compute Platform 2000 for
Microsoft® Private Cloud
Reference Architecture Guide
By Leo Nguyen and Rick Andersen
November 2015
Feedback
Hitachi Data Systems welcomes your feedback. Please share your thoughts by sending an email message to
[email protected] To assist the routing of this message, use the paper number in the subject and the title of this white
paper in the text.
Contents
Solution Overview................................................................................................................ 3
Key Solution Components.................................................................................................. 6
Hardware Components ................................................................................................................... 6
Brocade Switches ........................................................................................................................... 7
Software Components .................................................................................................................... 7
Solution Design..................................................................................................................... 9
Compute Architecture ..................................................................................................................... 9
Storage Architecture ..................................................................................................................... 10
Network Architecture .................................................................................................................... 13
Management Architecture............................................................................................................. 14
Engineering Validation ...................................................................................................... 18
Test Methodology ......................................................................................................................... 18
Test Results .................................................................................................................................. 18
1
Hitachi Unified Compute Platform 2000 for Microsoft® Private
Cloud
Reference Architecture Guide
Hitachi Unified Compute Platform 2000 for Microsoft® Private Cloud provides rapid provisioning and de-provisioning of
virtual machines. It also provides tight integration of the storage architecture. The solution has automation to provision a
new LUN to support additional virtual machines and add that LUN to a host cluster. This solution provides high availability
to management systems.
Unified Compute Platform 2000 for Microsoft Private Cloud is an entry level converged infrastructure. A converged
infrastructure is a single, optimized platform that combines compute, storage and networking. It brings tremendous
flexibility and scalability to manage, monitor and provision virtual machine resources in the data center. The converged
infrastructure is based on the following hardware components.

One, two, three, or four QuantaPlex T41S-2U compute chassis

One Hitachi Virtual Storage Platform G200

One Brocade ICX7450 switch

Two Brocade 6740 switches

Two Brocade 6505 switches (optional)
Unified Compute Platform for Microsoft Private Cloud supports the shift from server operator to a service provider. It has
a set of services to accompany the infrastructure, such as reporting, usage metering, and self-service provisioning.
Without these United Compute Platform services to accompany the infrastructure, the cloud service layer is unavailable.
Your investment would be little more than a traditional data center.
Unified Compute Platform 2000 for Microsoft Private Cloud provides the following:

A more cost-effective Unified Compute Platform system with a smaller size

A single converged infrastructure that runs VMware, Microsoft, and OpenStack virtualized workloads

A converged hardware stack for the following:

Remote office, branch office (ROBO)

Tier 2 workloads

Development or test environments

Virtual desktop infrastructure (VDI)
1
2
With
this solution you have the following:



Faster deployment with the ability to adapt to market pressures by leveraging the scalable design of the architecture.
Reduced risk with a tested and validated solution with end-to-end interoperability of compute, storage, and network
components.
Predictability with a consistent experience to the hosted workloads with standardized physical servers, network
devices, and storage systems.
This solution uses Hitachi Virtual Storage Platform G200. It scales as you host additional workloads in the cloud, providing
high availability to these workloads.
This reference architecture is for IT administrators involved in data center planning and design, particularly a Microsoft
Hyper-V® private cloud infrastructure. You need some familiarity with the following:

Hitachi Virtual Storage Platform Gx00 models

Hitachi Storage Navigator

Microsoft Windows Server® 2012 R2

Microsoft SQL Server® 2012

Microsoft System Center Virtual Machine Manager 2012 R2
Note — Testing of this configuration was in a lab environment. Many things affect production environments beyond
prediction or duplication in a lab environment. Follow the recommended practice of conducting proof-of-concept
testing for acceptable results in a non-production, isolated test environment that otherwise matches your production
environment before your production implementation of this solution.
2
3
Solution Overview
Hitachi Unified Compute Platform 2000 for Microsoft Private Cloud is built on hardware and software from Hitachi,
Quanta, Brocade, and Microsoft.


The QuantaPlex T41S-2U compute blade provides the compute resources.
Hitachi Virtual Storage Platform G200 provides the storage resources for Windows Server SAN boot, infrastructure,
management, and tenant virtual machines.

The Brocade ICX7450 switch provides the network management resources.

The Brocade VDX6740 switch provides local area network resources.
Unified Compute Platform 2000 for Microsoft Private Cloud is configured using a four node Hyper-V cluster that hosts a
mixed of infrastructure, management, and tenant virtual machines. The cluster contains the Hitachi and Microsoft
software required to deploy virtual machines to the cluster along with the products and tools to manage the Hyper-V
private cloud infrastructure components. This reference architecture provides the following capabilities:






Hyper-V host failover cluster — Failover clusters provide high availability and scalability to the server workloads.
Hyper-V virtual machine live migration — Moves running virtual machines from one physical server to another with
no impact on virtual machine availability to users.
Hype-V storage live migration — Moves running virtual machines storage from one cluster shared volume to
another with no impact on virtual machine availability to users.
Template based virtual machine provisioning — Virtual machine templates allow administrators to rapidly deploy
virtual machines.
Integration with System Center Operations Manager — Hitachi provides monitoring packs for Hitachi Virtual
Storage Platform Gx00.
Hitachi Storage Adapter for Microsoft Windows PowerShell — Allows an administrator to discover, manage, and
automate using scripts to automate complex tasks.
Figure 1 on page 4 provides a high-level design of the Unified Compute Platform 2000 for Microsoft Private Cloud base
unit. Figure 2 on page 5 provides a high-level design of the scale out of Unified Compute Platform 2000 for Microsoft
Private Cloud to a fully populated unit.
3
4
Figure 1
4
5
Figure 2
5
6
Key Solution Components
These are the key hardware and software components used to deploy Hitachi Unified Compute Platform 2000 for
Microsoft Private Cloud.
Hardware Components
Figure 1 lists the hardware components used in this solution.
Table 1. Hardware Components
Hardware
Description
Version
Quantity
QuantaPlex T41S-2U

Firmware Revision:
3.30.00
1




Hitachi Virtual Storage
Platform G200
SFF Disk Tray
Brocade VDX 6740 switch
Brocade ICX7450 switch
4-node chassis
2 Intel Xeon E5-2620 v3 processors,
BIOS Version:
2.40 GHz
S2S_3A14
192 GB memory
Emulex HBA
Intel 82599 10GigE OCP dual-port Firmware v2.01A10
card
Intel driver
Emulex LPe12002 dual-port 8GB
HBA card
3.9.58.9101

Dual controller
83-01-21-40/00
1

16 × 8 Gb/sec Fibre Channel ports
24 × 1.2 TB 10k RPM SAS disks
N/A
1
48-port 10 GbE switch
5.0.1d
2




Emulex HBA driver
v10.4.246.0
64 GB cache memory
24-port 1 GbE management switch 08.0.20c
1
Hitachi Virtual Storage Platform Gx00 (Midrange) Model Family
The Hitachi Virtual Storage Platform Gx00 model family systems are based on industry-leading enterprise storage
technology. With flash-optimized performance, these systems provide advanced capabilities previously available only in
high-end storage arrays. With the Virtual Storage Platform Gx00 family, you can build a high performance,
software-defined infrastructure to transform data into valuable information.
Hitachi Storage Virtualization Operating System provides storage virtualization, high availability, superior performance,
and advanced data protection for all models in the Virtual Storage Platform family. This proven, mature software provides
common features to consolidate assets, reclaim space, extend life, and reduce migration effort. New management
software improves ease of use to save time and reduce complexity. The infrastructure of Storage Virtualization Operating
System creates a management framework for improved IT response to business demands.
This solution uses Virtual Storage Platform G200.
6
7
Quanta
Cloud Technology QuantaPlex T41S-2U
Quanta Cloud Technology QuantaPlex T41S-2U is a rack mounted server designed for optimal performance and power
efficiency which supports up to 1.5 TB highly scalable memory capacity. It is powered by the Intel Xeon E5-2600 v3
processor product family for complex and demanding workloads. It supports flexible OCP and PCIe I/O expansion card
options.
Brocade Switches
Brocade and Hitachi Data Systems partner to deliver storage networking and data center solutions. These solutions
reduce complexity and cost, as well as enable virtualization and cloud computing to increase business agility.
The solution using the following Brocade products:

VDX 6740
Brocade VDX 6740 is a 1U 48-port rack mounted switch that delivers high performance and reduces network
congestion with 10 GbE ports, low latency, and deep buffers.

ICX 7450
Brocade ICX 7450 is a 1U 24-port rack mounted switch that delivers the performance, flexibility, and scalability
require for enterprise gigabit Ethernet access deployment. This switch is used for management networks only.

6505
Brocade 6505 is a 1U 24-port 8 Gb/sec Fibre Channel rack mounted switch that delivers reliable and high
performance storage area network. The basic configuration uses direct connect. When you scale out to three or four
compute chassis, then two Brocade 6505 switches are required.
Software Components
Table 2 lists the important software used in this solution.
Table 2. Software Components
Software
Version
Hitachi Storage Virtualization Operating System with Microcode Dependent
Hitachi Dynamic Provisioning
Hitachi Storage Navigator
Microcode Dependent
Hitachi Compute Advisor
1.1
Hitachi Infrastructure Director
1.1
Hitachi Storage Adapter for Microsoft® Windows
PowerShell
6.1.0
Hitachi Storage Adapter for Microsoft® System
Center Virtual Machine Manager
1.2.0
Microsoft Windows Server with Hyper-V
2012 R2 Standard and Datacenter edition
Microsoft SQL Server
2012 SP1 Enterprise edition
Microsoft System Center Virtual Machine Manager
2012 R2
7
8
Hitachi
Storage Virtualization Operating System
Hitachi Storage Virtualization Operating System spans and integrates multiple platforms. It is integrates storage system
software to provide system element management and advanced storage system functions. Used across multiple
platforms, Storage Virtualization Operating System includes storage virtualization, thin provisioning, storage service level
controls, dynamic provisioning, and performance instrumentation.
Storage Virtualization Operating System includes standards-based management software on a Hitachi Command Suite
base. This provides storage configuration and control capabilities for you.
Storage Virtualization Operating System uses Hitachi Dynamic Provisioning to provide wide striping and thin provisioning.
Dynamic Provisioning provides one or more wide-striping pools across many RAID groups. Each pool has one or more
dynamic provisioning virtual volumes (DP-VOLs) without initially allocating any physical space. Deploying Dynamic
Provisioning avoids the routine issue of hot spots that occur on logical devices (LDEVs).
Hitachi Compute Advisor
Hitachi Compute Advisor (HCA) is a software management tool used to configure, manage, and monitor server resources.
Hitachi Infrastructure Director
Hitachi Infrastructure Director (HID) is a software management tool used to provision and manage storage resources.
Hitachi Storage Adapter for Microsoft® Windows PowerShell
Hitachi Storage Adapter for Microsoft Windows PowerShell with the Hitachi Storage cmdlets allow an administrator to
discover, manage, and automate using scripts to automate complex tasks.
Hitachi Storage Adapter for Microsoft® System Center Virtual Machine Manager
Hitachi Storage Adapter for Microsoft System Center Virtual Machine Manager allows LUN provision to hosts and clone
virtual machine using Hitachi Snapshot or ShadowImage technology.
Microsoft Windows Server with Hyper-V
Microsoft Windows Server with Hyper-V is a virtualization platform that helps organizations of all sizes realize considerable
cost savings and operational efficiencies. With industry-leading size and scale, Hyper-V is the platform of choice for you
to run your mission critical workloads.
Microsoft SQL Server
Microsoft SQL Server is a complete set of enterprise-ready technologies and tools to derive the most value from
information at the lowest total-cost-of-ownership. Enjoy high levels of performance, availability, and security; employ
more productive management and development tools; and deliver pervasive insight with self-service business intelligence
(BI).
Microsoft Virtual Machine Manager
Virtual Machine Manager, a part of Microsoft System Center, is a management solution for the virtualized datacenter. It
lets you configure and manage your virtualization host, networking, and storage resources to create and deploy virtual
machines and services to private clouds
8
9
Solution Design
The detailed design for Hitachi Unified Compute Platform 2000 for Microsoft Private Cloud includes the following:

Compute Architecture
The basic solution uses one QuantaPlex T41S-2U compute chassis.


For a scale out configuration, use up to four QuantaPlex T41S-2U compute chassis
Storage Architecture
The basic solution uses one Hitachi Virtual Storage Platform G200, one SFF drive expansion tray with 24 x 1.2 TB 10K
SAS drive to support Windows Server SAN boot, infrastructure, management, and tenant virtual machines.


Switch Architecture



For a scale up configuration, use up to seven SFF drive expansion trays. Each tray houses 24 × 2.5 inch disks.
The basic solution uses uses one Brocade ICX 7450 switch for management, two Brocade VDX 6740 switches for
local area network, and direct connect from compute HBA ports to storage ports.
For a scale out configuration to three or four compute chassis, use two Brocade 6505 switches for the storage
area network.
Management Architecture
The management architecture uses the following:

Hitachi Compute Advisor

Hitachi Infrastructure Director

Microsoft System Center Virtual Machine Manager
Compute Architecture
Hitachi Unified Compute Platform 2000 for Microsoft Private Cloud uses one QuantaPlex T41S-2U compute chassis. It
can scale out to support two, three or four chassis.
Each chassis supports up to four compute blades.
Compute Chassis
QuantaPlex T41S-2U compute chassis consists of four mainboard sleds and two power supply units. The power supply
units are directly connected to the power distribution boards, the HDD backplane, and middle plane. This allows each
power supply unit to individually provide power for all mainboards (MB1 to MB4).
9
10
Compute
Blade
The compute blade architecture is a critical component of the virtualized infrastructure. Powered by the Intel Xeon
E5-2600 v3 processor product family and DDR4 memory technology, the compute blade provides flexible I/O scalability.
Each compute blade uses the following hardware components.

2 Intel Xeon E5-2620 v3 processor at 2.40 GHz

192 GB memory

Intel 10 GbE network adapter

Emulex 8 GB/sec Fibre Channel host bust adapter
Figure 3 shows the QuantaPlex T41S-2U rear view I/O ports for blade 1 to blade 4.
Figure 3
Storage Architecture
Hitachi Unified Compute Platform 2000 for the Microsoft Private Cloud uses the following on Hitachi Virtual Storage
Platform G200:

One controller

One service processor (SVP)

One SFF disk expansion tray with 24 × 1.2 TB 10k RPM SAS disks.
On Virtual Storage Platform G200, Hitachi Dynamic Provisioning, a part of Hitachi Storage Virtualization Operating
System, provisions the storage pools. Using Hitachi Dynamic Provisioning provides the following benefits for you:


Wide striping — Dynamic Provisioning avoids routine issue of hot spots that occur on logical devices (LDEVs). These
occur within individual RAID groups when the host workload exceeds the IOPS or throughput capacity of that RAID
group. Dynamic provisioning distributes the host workload across many RAID groups, which provides a smoothing
effect that dramatically reduces hot spots.
Thin provisioning functionalities — Physical space assignment from the pool to the dynamic provisioning volume
happens as needed using 42 MB pages, up to the logical size specified for each dynamic provisioning volume. There
can be a dynamic expansion or reduction of pool capacity without disruption or downtime. You can rebalance an
expanded pool across the current and newly added RAID groups for an even striping of the data and the workload.
10
11
Direct
Connect Configuration
The solution uses two 8 Gb/sec 2-port Emulex Fibre Channel card installed on the PCIe slot of the compute blade as
shown in Figure 4.
Figure 4
Each server blade uses the ports in Table 3.
Table 3. Server Blade Port Configuration
Compute Chassis Number
Server Blade Number
Port
1
1
1A/2A
2
3A/4A
3
5A/6A
4
7A/8A
Configure the storage port property values as shown in Table 4. On the Emulex host bus adapter BIOS, configure the
connection type as FC-AL.
Table 4. Storage Port Properties
Property
Value
Port Attribute
Target
Port Security
Disable
Port Speed
Auto (8 Gb/sec)
Fabric
OFF
Connection Type
FC-AL
Use Microsoft Multipath I/O for multipathing, using the round-robin policy. Round-robin load balancing in Multipath I/O
automatically selects a path by rotating through all available paths. By balancing the load across all available paths, you
optimize IOPS and response time.
11
12
RAID
Configuration
This solution uses 16 × 1.2 TB 10k RPM SAS disks. Create the parity groups as follows:

One parity group for RAID-10 (2D+2D), using four disks

Two parity groups for RAID-6 (6D+2P), using 12 disks.
Dynamic pools can be expanded by adding additional parity groups to support additional work loads and virtual
machines.
Pool Configuration
Create three dynamic provisioning pools, as shown in Table 5.
Table 5. Dynamic Provisioning Pool Configuration
Dynamic
Provisionin
g Pool
RAID
Configuration
Number
of Parity
Groups
Number
of Disks
Capacity
Purpose
0
RAID-10 (2D+2D)
1
4
2 TB
1
RAID-6 (6D+2P)
1
8
3 TB


2
RAID-6 (6D+2P)
1
8
6 TB

SAN operating system boot
Infrastructure and
management virtual
machines
Tenant virtual machines
Clustered Shared Volumes Configuration
Hitachi Unified Compute Platform 2000 for Microsoft Private Cloud implements clustered shared volumes (CSV) to host
the virtual machine operating system and application data. CSVs are exclusively for use with Microsoft Hyper-V failover
clustering. They enable all nodes in the cluster to access the same cluster storage volumes at the same time.
Using CSVs eliminates the one virtual machine per LUN requirement, allowing multiple virtual machines to be placed on a
single CSV. This simplifies the management of the storage infrastructure in a private cloud environment.
Because all cluster nodes can access all CSVs simultaneously, standard LUN allocation methodologies based on
performance and capacity requirements of the expected workloads can be used.

Pool 0 contains the following:


Pool 1 contains the following:


SAN operating system boot volumes for Quanta servers
Infrastructure and management virtual machines
Pool 2 contains the following:

Tenant virtual machines
CSV architecture differs from other traditional clustered file systems. This frees it of scalability limitations, such as one
virtual machine per LUN and drive letter limitations. As a result, there is no special guidance for scaling the number of
Hyper-V nodes or virtual machines on a CSV volume.
12
13 virtual disks of all virtual machines running on a particular CSV contend for storage I/O. To best implement your
The
solution, do the following:


Understand the I/O workload characteristics for the virtual machines hosted on the CSVs located in dynamic
provisioning pools.
Take into consideration the IOPS requirements of any virtual machine to be deployed along with its I/O profile. For
example, consider the requirements of random read and write operations versus sequential write operations.
Network Architecture
Hitachi Unified Compute Platform 2000 for Microsoft Private Cloud uses two Brocade VDX 6740 switches for network
traffic. Configure the two Brocade VDX 6740 switches as link aggregation using Brocade Virtual Cluster Switching and
Brocade Inter-Switch Link technologies for redundancy and performance. Each switch has 48 ports.
For performance enhancement and security, configure the networks using the following VLANs:

Management Network
The switches provide connectivity between the Microsoft Hyper-V hosts and Microsoft Virtual Machine Management.
It is recommended to leave management network as untagged.

Cluster Network
VLAN 411provides inter-node cluster communication. This includes the cluster heartbeat and cluster shared volumes
(CSV) redirection.

Live Migration Network
VLAN 412 provides virtual machine live migration.

Virtual Machine Network
VLAN 413 provides virtual machine connectivity.
Each QuantaPlex compute blade uses the following:

An onboard 1 GbE adapter for management

A dual-port Intel 10 GbE OCP card for network traffic
Use Virtual Machine Manager in Microsoft System Center to configure the logical network, logical switches, and port
profiles. Configure the network adapters as follows for NIC teaming:

Team mode — Switch independent

Load balancing mode — Dynamic

Standby adapter — None (all adapter active)
13
14
Figure
5 shows the rear view of the QuantaPlex compute chassis connected to the Brocade VDX 6740 switches for
network traffic.
Figure 5
Management Architecture
Managing the infrastructure for Hitachi Unified Compute Platform 2000 for Microsoft Private Cloud is critical to the health
and operation of the compute, storage, and networking. Use these software tools for management:

“Hitachi Compute Advisor” on page 14

“Hitachi Infrastructure Director ” on page 15

“Hitachi Storage Adapter for Microsoft Windows PowerShell” on page 15

“Hitachi Storage Adapter for Microsoft System Center Virtual Machine Manager” on page 15

“Microsoft System Center 2012 R2 Virtual Machine Manager ” on page 15
Hitachi Compute Advisor
Use Hitachi Compute Advisor to do the follow tasks:



Power management options

Turn on the power of a managed server and start the operating system.

Shut down the operating system and turn off the power of a managed server.
Firmware management options

Update compute BIOS

Update compute BMC
Monitor managed servers

SNMP trap reception settings

SNMP trap forward settings
14
15
Hitachi
Infrastructure Director
Use Infrastructure Director to create, edit, view, and delete the following:

Parity groups

Pools

Volumes
Hitachi Storage Adapter for Microsoft Windows PowerShell
Use Hitachi Storage Adapter for Microsoft Windows PowerShell to automate complex tasks.
Hitachi Storage Adapter for Microsoft System Center Virtual Machine Manager
Use Hitachi Storage Adapter for Microsoft System Center Virtual Machine Manger to provision LUN to hosts and clone
virtual machines using Hitachi Snapshot or ShadowImage technology. A license is required to enable Snapshot or
ShadowImage on the storage.
Microsoft System Center 2012 R2 Virtual Machine Manager
Use Microsoft System Center Virtual Machine Manager to install, configure, deploy, and maintain Microsoft Hyper-V
hosts, virtual machines, and virtual network.
Microsoft Hyper-V Host Installation and Configuration
Hitachi Unified Compute Platform 2000 for Microsoft Private Cloud uses a base configuration of four Hyper-V hosts. It can
scale out to support 16 nodes if needed.
Install and configure Microsoft Windows Server 2012 R2. Present the volumes to all Hyper-V hosts shown in Table 6.
Install and configure Windows multipathing software for multipath I/O.
Table 6. Volume Configuration
Hyper-V Number
Pool Number
Pool Volume
Capacity
Purpose
1
0
0
1 GB
Quorum drive for Hyper-V cluster
1
1
2 TB
Infrastructure and management
cluster shared volume
2
2
4 TB
Tenant cluster shared volume
2
3
4
1
2
3
4
1
2
3
4
15
16
Virtual
Machine Installation and Configuration
Hitachi Unified Compute Platform 2000 for Microsoft Private Cloud requires the infrastructure and management virtual
machines shown in Table 7.
Table 7. Required Infrastructure and Management Virtual Machine
Number of Virtual
Machines Required
Purpose
Virtual Machine Names
1
Domain controllers
DC1, DC2
2
Hitachi Compute Advisor
HCA
2
Hitachi Infrastructure Director
HID
2 (2 recommended)
Microsoft SQL Server
SQL
2
Microsoft System Center with Virtual Machine
Manager
SCVMM
Configure the vCPU, vRAM, vDISK and vNIC for the virtual machines shown in Table 8. Create the VHDX disks as fixed
disk for best performance.
In a production environment, it is recommended to use the following:

If it exists, join an existing Microsoft Active Directory® domain. Otherwise create a new Active Directory domain.

Create two SQL Server virtual machines for redundancy.
Table 8. Virtual Machine Configuration
Hyper
-V
Host
Cluster
Shared
Volume
Virtual
Machine
Name
vCPU
vRAM
(GB)
vDisk
vNIC
Purpose
0
0
HCA
4
8
100
1
Compute management
1
HID
4
8
Storage management
1
DC1
2
4
Primary domain controller for
Active Directory
2
DC2
2
4
Secondary domain controller
for Active Directory
3
SCVMM
2
24
Virtual Machine Manager
4
SQL
2
24
SQL Server database for
SCVMM
16
17
Microsoft
Hyper-V Cluster Configuration
Add all four Microsoft Hyper-V hosts for management to Virtual Machine Manager in Microsoft System Center. Create a
Hyper-V cluster. During the cluster creation process, Virtual Machine Manager does the following:

Validates that all hosts meet the prerequisites, such as required operating system and domain membership

Enables the failover clustering on each host

Unmasks the selected storage logical units to each host

Creates the configured external virtual networks

Runs the cluster validation process

Creates the cluster with quorum and enables cluster shared volumes

For each logical unit that is designated as a cluster shared volume, assigns the logical unit as a cluster shared volume
on the cluster
Virtual Network Configuration
In keeping with Microsoft best-practice recommendations for private cloud implementations, do the following:

Break network traffic down into separate networks.

Assign each network type to a different subnet.

Use sub-netting to break the configuration into smaller, more efficient networks.

Achieve further isolation of network types by using VLAN isolation and dedicated network switches.
For VLAN-based network segmentation or isolation, several components must be configured specifically to enable rapid
provisioning and network segmentation.
With Microsoft Hyper-V and host clusters, define identical virtual networks on all nodes. This is necessary so a virtual
machine is able to failover to any node and to maintain its connection to the network.
17
18
Engineering Validation
This describes the tools, test methodology, and test results used to validate this solution for Hitachi Unified Compute
Platform 2000 for the Microsoft Private Cloud.
Test Methodology
This test methodology validated the storage disk subsystem.
The purpose of using IOmeter is to validate the maximum IOPS the storage subsystem can produce using a RAID-6
(4D+2P) configuration for 6 × 1.2 TB 10k RPM SAS disks with a latency of less than 20 msec.
Table 9 shows the IOmeter parameters configuration.
Table 9. IOmeter Configuration
Property
Value
Number of worker
2
Maximum disk size
80%
Number of outstanding I/Os
32
Assigned access specifications Web server
Run time
1 minute
Number of Run
8
Total run time
8 minutes
Test Results
These are the test results from IOmeter.
These tables show the IOmeter test results for 4 KB (Table 10 on page 19) and 8 KB (Table 11 on page 19) using a RAID-6
(4D+2P) configuration for 6 × 1.2 TB 10k RPM SAS disks with a latency of less than 20 msec. These results provide a
baseline used to determine the number of virtual machines to support a specific level of IOPS.
Testing for this enviornment used the following standard Microsoft profile:


A minimum of 50 IOPS per virtual machine
Up to a maximum of 100 IOPS per machines when deploying virtual machines in a private cloud with a 4KB I/O size
available to each virtual machine.
For some workloads this may be too little and for other workloads too large. Hitachi Data Systems and you should work
together to determine appropriate sizing and solution hardware alignment.
18
19
Table 10. IOmeter test results for 4 KB
Property
Value
Number of Disk
RAID Level
Size (KB)
4096
6
RAID 6 (4D+2P)
Total IOPS
1209
Average Response Time (msec)
13.23
Number of supported VM total IOPS (50 24
IOPS per VM)
Number of supported VM total IOPS
(100 IOPS per VM)
12
Table 11. IOmeter test results for 8K
Property
Value
Number of Disk
RAID Level
Size (KB)
8192
6
RAID 6 (4D+2P)
Total IOPS
1728
Average Response Time
18.5
Number of supported VM total IOPS
(50 IOPS per VM)
34
Number of supported VM total IOPS
(100 IOPS per VM)
17
19
For More Information
Hitachi Data Systems Global Services offers experienced storage consultants, proven methodologies and a
comprehensive services portfolio to assist you in implementing Hitachi products and solutions in your environment. For
more information, see the Hitachi Data Systems Global Services website.
Live and recorded product demonstrations are available for many Hitachi products. To schedule a live demonstration,
contact a sales representative. To view a recorded demonstration, see the Hitachi Data Systems Corporate Resources
website. Click the Product Demos tab for a list of available recorded demonstrations.
Hitachi Data Systems Academy provides best-in-class training on Hitachi products, technology, solutions and
certifications. Hitachi Data Systems Academy delivers on-demand web-based training (WBT), classroom-based
instructor-led training (ILT) and virtual instructor-led training (vILT) courses. For more information, see the Hitachi Data
Systems Services Education website.
For more information about Hitachi products and services, contact your sales representative or channel partner or visit
the Hitachi Data Systems website.
1
Corporate Headquarters
2845 Lafayette Street
Santa Clara, CA 96050-2639 USA
www.HDS.com
community.HDS.com
Regional Contact Information
Americas: +1 408 970 1000 or [email protected]
Europe, Middle East and Africa: +44 (0) 1753 618000 or [email protected]
Asia Pacific: +852 3189 7900 or [email protected]
HITACHI is a trademark or registered trademark of Hitachi, Ltd. © Hitachi Data Systems Corporation 2015. All rights reserved. Microsoft, Hyper-V, Windows Server, SQL Server, and
Active Directory are trademarks or registered trademarks of Microsoft Corporation. All other trademarks, service marks, and company names are properties of their respective
owners.
Notice: This document is for informational purposes only, and does not set forth any warranty, expressed or implied, concerning any equipment or service offered or to be offered by
Hitachi Data Systems Corporation.
AS-448-00. November 2015.
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Related manuals

Download PDF

advertisement