HPE Reference Architecture for Microsoft SQL Server 2016 on HPE

HPE Reference Architecture for
Microsoft SQL Server 2016 on HPE
Hyper Converged 250 for Microsoft
Cloud Platform System Standard
Technical white paper
Technical white paper
Contents
Executive summary .................................................................................................................................................................................................................................................................................................................................3
Introduction ....................................................................................................................................................................................................................................................................................................................................................3
HPE Hyper Converged 250 family .....................................................................................................................................................................................................................................................................................3
Solution overview ..................................................................................................................................................................................................................................................................................................................................... 4
Solution components............................................................................................................................................................................................................................................................................................................................ 5
Hardware................................................................................................................................................................................................................................................................................................................................................... 6
Software ..................................................................................................................................................................................................................................................................................................................................................... 7
Application software ....................................................................................................................................................................................................................................................................................................................... 7
Best practices and configuration guidance for the solution ............................................................................................................................................................................................................................. 7
Capacity and sizing ................................................................................................................................................................................................................................................................................................................................ 9
Scale-up scenario .............................................................................................................................................................................................................................................................................................................................. 9
Scale-out scenario...........................................................................................................................................................................................................................................................................................................................10
Consolidation scenario.................................................................................................................................................................................................................................................................................................................11
Analysis and recommendations .........................................................................................................................................................................................................................................................................................12
Summary ........................................................................................................................................................................................................................................................................................................................................................ 13
Implementing a proof-of-concept .................................................................................................................................................................................................................................................................................... 13
Appendix A: Bill of materials ........................................................................................................................................................................................................................................................................................................ 13
Resources and additional links ..................................................................................................................................................................................................................................................................................................14
Technical white paper
Page 3
Executive summary
As a business continues to grow and evolve so will their IT environment. To accelerate growth in the business engine, there is a need for an agile,
flexible and scalable IT environment to support it. In a fast growing business, enterprise architects, database architects and administrators are
tasked with supporting diverse application workloads, delivering greater performance, capacity and easy scalability, at lower cost. Adding more
traditional hardware silos into the data center to meet the ever growing demands may not be viable. To thrive in an ever demanding idea
economy, IT needs to be scalable as workload demands increase, allowing businesses to respond in days not weeks and months.
Hewlett Packard Enterprise Hyper Converged 250 (HC 250) solutions make excellent choices for providing turnkey hyper-converged,
virtualization solutions for medium-sized businesses and enterprise remote offices with a choice of VMware® vSphere or Microsoft® Hyper-V.
These solutions make it easy for a business to start small and grow on-demand by providing a cost effective, simple, agile and highly available
platform. HPE HC 250 forms a basic building block, simplifying the delivery of virtualized servers, storage and networking, thus enabling a
business to scale on an as needed basis, adding one block at a time as the business grows.
This paper demonstrates the performance and scalability of Microsoft SQL Server 2016 on HPE HC 250 for Microsoft Cloud Platform System
(CPS) Standard. The HPE HC 250 for Microsoft CPS Standard provides an excellent platform for quick and efficient deployment of small and
medium virtualized databases. The integration with Microsoft System Center console makes it easy to use and simple to manage. This paper
provides details on configuration best practices for SQL Server database deployments. The white paper also presents the results of scale-up,
scale-out and database consolidation tests performed using an OLTP database workload running against the SQL Server instances.
Target audience: Information in this paper will be useful for IT generalists, data center managers, enterprise architects, deployment /
implementation engineers, Microsoft SQL Server engineers, DBAs and others wishing to learn more about SQL Server instances running on
virtual machines on HPE HC 250. A working knowledge of Microsoft SQL Server, server architecture, networking and storage is recommended.
This white paper describes testing completed in March 2016.
Document purpose: The purpose of this document is to describe performance and scalability of SQL Server on virtual machines hosted on the
HPE hyper-converged appliance, HPE HC 250, highlighting recognizable benefits to technical audiences.
Introduction
This reference architecture demonstrates the use of the HPE HC 250 for Microsoft CPS Standard to deploy Microsoft SQL Server 2016. The focus
of this paper is to show how the HPE HC 250 for Microsoft CPS Standard can scale-up as well as scale-out to meet the demands of SQL Server
database growth and increasing workload.
Deploying Microsoft SQL Server on the HPE HC 250 for Microsoft CPS Standard has the following benefits.
• Rapidly develop and deliver new services for business by easily and quickly deploying virtual infrastructure.
• Reduce infrastructure and management complexity by consolidating multiple databases on a single platform.
• Scale business services and virtual infrastructure on-demand by adding more HPE HC 250 for Microsoft CPS Standard appliances.
• Protect your services with built-in data protection features and disaster recovery capabilities.
• Easy access and integration with Microsoft Azure services.
HPE Hyper Converged 250 family
The HPE HC 250 solution offers a simple, agile, and highly available virtualization appliance that provides an optimized solution set of HPE
servers, HPE StoreVirtual VSA shared storage, and leading virtualization software that is configurable and ready for VM provisioning. The HPE
HC 250 family supports VMware and Microsoft virtualization technologies, and continues to provide a turn-key virtualization solution for
medium-sized businesses, enterprises, and IaaS providers.
Designed from the ground up for the software-defined data center, the HPE HC 250 solution enables a standardized appliance approach to
virtual machine deployment. This is regardless of whether the system is used as a primary virtualization platform in medium-sized businesses or
as a dedicated resource pool for specific applications in the enterprise. Unlike other hyper-converged systems on the market, the HPE HC 250
solutions can be customized and are ready for virtualized workloads in a few mouse clicks.
Technical white paper
Page 4
All hardware and software components are pre-installed and pre-integrated by Hewlett Packard Enterprise. Unique to the HPE HC 250 family,
quick customization using HPE OneView InstantOn software enables faster time to value. After the initial installation, IT administrators can
manage their physical and virtualized environment with HPE OneView for VMware vCenter, or with HPE OneView for Microsoft System Center.
Solution overview
Working together, Hewlett Packard Enterprise and Microsoft developed an all-new, highly efficient, and flexible hyper-converged system called
the HPE Hyper Converged 250 for Microsoft Cloud Platform System (CPS) Standard. This factory-integrated, modular appliance is easy to use,
hybrid cloud-ready, scalable, and managed from within the Microsoft System Center console. The Microsoft software that makes up the CPS
Standard offering is also pre-installed and configured at the factory.
Figure 1. HPE HC 250 system
The HC 250 for Microsoft CPS Standard system is pre-integrated with Microsoft Windows Server® 2012 R2 Datacenter Edition as the base
operating system, Hyper-V for the hypervisor, System Center Operations Manager, System Center Virtual Machine Manager (VMM) for
management, HPE StoreVirtual VSA for clustered storage and Windows® Azure Pack with WAP Tenant Portal and WAP Admin Portal for selfservice, multi-tenant cloud services. The HPE OneView for Microsoft System Center plug-in enables management of HC 250 for Microsoft CPS
Standard from the System Center VMM console, providing a single management window for day-to-day management activities across physical,
virtual, and cloud environments. The Microsoft Cloud Platform System Standard software also provides:
• Azure-consistent cloud-in-a-box with add-on subscriptions to Azure-connected services such as Azure Backup and Site Recovery.
• Familiar and consistent end-user experience through the Windows Azure Pack.
HPE StoreVirtual VSA plays an important role in managing storage across all nodes of the HC 250 for Microsoft CPS Standard system. An
instance of VSA runs as a virtual machine on each node of the HC 250 for Microsoft CPS Standard, making all the local disks into a shared pool of
storage. HPE StoreVirtual VSA Software uses scale-out, distributed clustering to provide a pool of storage with enterprise storage features and
simple management at a reduced cost. Multiple HPE StoreVirtual VSAs running on multiple servers create a clustered pool of storage with the
ability to make data highly available by protecting volumes with Network RAID. Adding more HPE StoreVirtual VSAs to the cluster grows the
storage pool. With Network RAID, blocks of data are striped and mirrored across multiple HPE StoreVirtual VSAs, allowing volumes and
applications to stay online in the event of disk, storage subsystem, or server failure. iSCSI connectivity on HPE StoreVirtual VSA supports the use
of the storage pools by hypervisors.
The solution described in this paper used a factory installed image of the base operating system. Once the system is racked and turned on, an
administrator can log in to the management VM and launch HPE OneView InstantOn to complete the deployment. OneView InstantOn is the
configuration tool that is preinstalled on the system. This tool guides you through the steps to get the system ready for VM provisioning. The
following figure shows a high-level block diagram of the solution configuration that was tested.
Technical white paper
Page 5
Figure 2. Logical solution diagram of HC 250 with Microsoft SQL Server VMs
The system is pre-installed with HPE StoreVirtual VSA VMs and a management VM installed with tools for managing the system as well as
virtual machines. Using these tools, SQL Server VMs were created and placed on desired server nodes. Three separate networks for
management, storage and Microsoft SQL Server VMs were defined to isolate traffic as shown in figure 2. An external load generator server was
used to drive the load. It was connected on the same network as the SQL Server VMs.
Solution components
For this solution, Microsoft SQL Server 2016 instances were installed on virtual machines hosted on the HC 250 for Microsoft CPS Standard
system. SQL Server 2016 provides mission critical capabilities, scalability, performance and availability for OLTP environments. This white paper
provides scalability and performance test results of SQL Server 2016 on the HC 250 platform. While the testing focuses on SQL Server 2016, the
results are applicable for older SQL Server instances that can be migrated and consolidated to run on the HC 250 for Microsoft CPS Standard
platform.
The sections below describe both the hardware and software components that were used for testing.
Technical white paper
Page 6
Hardware
The HC 250 for Microsoft CPS Standard system is based on HPE Apollo Gen9 technology. In a 2U chassis there can be up to four individuallyserviceable servers. This solution used one HC 250 for Microsoft CPS Standard system with four server nodes in a single chassis. Figure 3 below
shows the four server nodes numbered 1 to 4 in a 2U form factor chassis. All server nodes have an identical configuration:
• Two Intel® Xeon® E5-2680 v3 processors @ 2.50 GHz with 12 cores each for a total of 24 cores per server
• 512 GB memory
• Two 10 GbE SFP+ network connectivity ports
• Hybrid storage configuration with four 1.2 TB HDD and two 480 GB SSDs
Figure 3. Rear view showing 4 nodes
Figure 4. Front view
As shown in figure 4, the system houses 24 disk drives. Each server node is connected to 6 of them – 2 SSDs and 4 HDDs. Each node has a local
RAID configuration for the disks attached to that server. HPE StoreVirtual VSA then creates a network RAID over these disks across all 4 server
nodes. This hybrid storage configuration gives approximately 13.7 TB of highly available storage for a four-node system. For network
connectivity, there are two 10 GbE ports on each server node which are teamed by the operating system. Each server node also has a dedicated
HPE Integrated Lights-Out (iLO) management port.
The solution also made use of the HPE 5920AF-24XG Switch for networking between the management, storage and SQL Server VM network.
The HPE 5920 Switch Series is high-density 10 GbE, ultra-deep packet buffering top-of-rack (ToR) switches. These switches are part of the HPE
FlexNetwork architecture's HPE FlexFabric solution module and are ideally suited for deployments at the server access layer of large enterprise
data centers. With the increase in virtualized applications and server-to-server traffic, businesses now require ToR switch innovations that will
meet needs for higher-performance server connectivity, convergence of Ethernet and storage traffic, the capability to handle virtual
environments, and ultra-deep packet buffering all in a single device.
Technical white paper
Page 7
Figure 5. HPE 5920AF-24XG Switch
Software
All four servers in the HPE HC 250 for Microsoft CPS Standard system are pre-installed with Microsoft Windows Server 2012 R2 Datacenter and
have integrated data services from HPE StoreVirtual VSA. The system is pre-integrated with Microsoft Hyper-V, System Center, System Center
Operations Manager and System Center Virtual Machine Manager. The system provides a virtualized, multi-tenant infrastructure based on
Hyper-V with a single pane-of-glass management using HPE OneView for Microsoft System Center. Windows Azure Pack is also factory-installed
making the system hybrid cloud-ready. Valid Microsoft Windows Server 2012 R2 Datacenter and System Center 2012 R2 licenses must be
procured by the customer.
HPE StoreVirtual technology on top of the Microsoft virtualization platform provides storage capability with superior high-availability and
disaster recovery capabilities. Each server on the HC 250 for Microsoft CPS Standard system has an HPE StoreVirtual VSA virtual machine
running, providing a 4-node clustered storage with:
• Adaptive Optimization for workload acceleration (only with Hybrid Storage option)
• Network RAID 0, 5, 6, 10, 10+1, and 10+2
• Integrated Thin Provisioning
• Virtual machine and application-consistent snapshots
• Multi-Site HA (Synchronous Replication across several locations)
• Remote Copy (Snapshot-based, asynchronous Replication with Bandwidth Throttling)
• Storage federation with Hewlett Packard Enterprise Storage products
Application software
The SQL Server version used for this testing was Microsoft SQL Server 2016 (CTP3.2) - 13.0.900.73 (X64). It was installed on virtual machines
running Windows Server 2016 Technical Preview 4.
Best practices and configuration guidance for the solution
The HC 250 for Microsoft CPS Standard system requires a minimum of two different networks for deployment – management and storage. For
this testing, a third network was also used on which SQL Server virtual machines were connected. To be able to successfully deploy the system,
the following requirements are mandatory:
• Management and storage networks must be routable.
• No VLAN tagging on the management network.
• Storage network must be a tagged VLAN to avoid collisions with management traffic. The third network for SQL Servers must also be a
tagged VLAN.
• Existing DNS and AD DS domains running at a Windows Server 2012 R2 Datacenter functional level.
Technical white paper
Page 8
Note
For details on how to set up and configure HC 250, please follow the Hyper Converged 250 for Microsoft Cloud Platform System Standard
Installation Guide.
Three types of scenarios were tested on the HC 250 for Microsoft CPS Standard system – scale-up, scale-out and database consolidation. The
goals of the testing were as follows:
• Scale-up – Measure SQL Server database performance scalability when more CPU and memory were added to a single VM.
• Scale-out – Determine how the system performance scales when running one SQL Server instance on each node of the HPE HC 250 for
Microsoft CPS Standard.
• Database consolidation – Showcase the performance of the system running a mixture of SQL Server instances with small and medium sized
databases on virtual machines in a consolidation scenario.
For all test cases, the configurations below were modified on the SQL Server VM:
• Allow Instant File Initialization (service account given “Perform Volume Maintenance Tasks”)
• To optimize for a strictly OLTP workload, set Maximum Degree of Parallelism to 1
• “Lock pages in memory” right assigned to user running SQL Server service
• Set Maximum worker threads to 3000
• Set Max Server Memory to 75% of allocated virtual memory of the virtual server
• The following trace flags were enabled:
– T1118 (improve tempdb concurrency)
– T1117 (reduce allocation contention in tempdb)
– T8048 (reduce CMEMTHREAD waits)
– T834 (use large page allocations in Windows)
– –k (set checkpoint throughput bandwidth limit)
Note
For further information about the –k parameter, please review the knowledge base article:
https://msdn.microsoft.com/en-us/library/ms189573.aspx
The following BIOS settings were enabled on all four server nodes of the HC 250 system:
• Hyper-Threading – Enabled
• Intel Turbo Boost – Enabled
• HPE Power Profile – Maximum Performance
For volumes created on VSA:
• All volumes used were in Network RAID-10.
• Adaptive Optimization (AO) was enabled on volumes hosting SQL Server data files. Table 1 below shows details of VSA volumes and AO
status.
• Volumes created were fully provisioned.
Technical white paper
Page 9
The virtual machines running SQL Server were created with multiple virtual disks with designated use as shown in table 1 below. Additionally:
• The virtual disks of the virtual machines that were used to store SQL Server data files were of fixed type.
• The virtual disks used for operating system and SQL binaries were of dynamic type.
• Multiple volumes were provisioned on VSA storage. The drives for the operating system and SQL binaries were placed on the same volume
and the rest of the drives were placed on separate volumes on the VSA as shown in the table below. This enabled isolation of the sequential
write workload on log files from the random read/write workload on data files.
Table 1. SQL Server VM drive configuration
DRIVE
USED FOR
VSA VOLUME NAME
ADAPTIVE OPTIMIZATION
C
Operating System
VM_SQL_OS
Disabled
E
SQL home
VM_SQL_OS
Disabled
F
Data files
DataVol
Enabled
G
Temp DB
TEMPDB
Enabled
H
Transaction log
LOGS
Enabled
As a best practice, virtual disks must be configured across multiple virtual SCSI adapters on the virtual machine for better performance. For this
test, three SCSI adapters were used. Drives F, G and H were placed under the first, second and third SCSI adapters respectively. Drives C and E
were configured as IDE devices and placed under an IDE adapter.
Capacity and sizing
An OLTP workload was used to evaluate all the test cases – scale-up, scale-out and consolidation – using the Microsoft BenchCraft load
generation tool. The BenchCraft test tool was installed on an external system in the lab. It was used to build the test databases and drive a
workload against those databases. Two database sizes, approximately 55 GB (small) and 100 GB (medium), were used. Databases were created
with enough disk capacity to accommodate their growing size when running tests. Tests were run for long durations to ensure frequently
accessed data blocks were migrated to the SSD tier by the AO feature on VSA. For this testing, up to four databases were consolidated under
one SQL Server instance giving a combined database size ranging from 55 GB to 265 GB. Before each test run, all databases were restored from a
backup, which was created immediately after the database was created.
Scale-up scenario
The purpose of this test case was to determine how SQL Server scales up when CPU and memory resources were added incrementally to the
SQL Server VM running on the HC 250 for Microsoft CPS Standard. An individual SQL Server VM can scale-up to consume the available
resources on any individual compute node in the system if there is a demanding SQL Server workload. Table 2 highlights the four tested VM
configurations in the scale-up testing.
Table 2 shows the four SQL Server VM configurations and the database size used for this use case.
Table 2. Scale-up VM sizes
VM CORE/MEMORY
DATABASE SIZE
4 vCPU / 16 GB
55 GB
8 vCPU / 32 GB
110 GB
16 vCPU / 64 GB
165 GB
24 vCPU / 96 GB
265 GB
Technical white paper
Page 10
During testing, CPU usage and disk latency thresholds were monitored. The load on the SQL Server was gradually increased until CPU utilization
was around 90% or disk latency exceeded a 10 ms threshold, thus ensuring there was no performance hindrance to the applications connecting
to the SQL Server databases. Using BenchCraft, a wide range of workloads were tested on different scale-up VM configurations. For the smallest
virtual machine configuration, a workload of 216,000 transactions per minute was driven on the SQL Server and for the largest configuration,
720,000 transactions per minute was used. The –k parameter was set to 78 to keep the disk latency under acceptable limits. This parameter
defines the checkpoint throughput bandwidth limit for the SQL Server. The value for –k depends on multiple factors. It depends on the load on
SQL Server, SQL Server buffer pool size and performance of disk subsystem. During higher loads, it was observed that SQL Server checkpoint
operations resulted in disk latencies above 10 ms. To reduce the latencies it was required to control the throughput bandwidth during checkpoint
operations using the –k parameter.
The graph in figure 6 shows the performance results obtained in terms of batch requests per second on a relative scale. The graph depicts the
incremental gains in performance as the SQL Server virtual machine resources increased. The ability for the HC 250 for Microsoft CPS Standard
to support a range of SQL Server workloads, from small to large, is critical in ensuring that the platform is flexible enough to meet the wide range
of performance requirements for SQL Server database workloads.
Scale-up
400
% Relative Performance
350
300
250
200
150
100
50
0
4 vCPU
16 GB
8 vCPU
32 GB
16 vCPU
64 GB
24 vCPU
96 GB
VM size
Figure 6. Scale-up performance
Scale-out scenario
This test case demonstrates the performance of the HC 250 for Microsoft CPS Standard system as SQL Server instances are added in a scale-out
fashion, exploring the possibility of running multiple VMs on the system without compromising the required performance needs for any
individual SQL Server instance. For this test case, a VM configuration with 4 cores and 48 GB memory was chosen with a database of size 110 GB.
Four SQL Server VMs were created, one on each server node. Testing was started with one SQL Server VM running on one server node first and
results were recorded. Next, a second SQL Server VM was started on the second node and tests were run in parallel on the two VMs. Then, with
three SQL Server VMs and finally with four SQL Server VMs running in parallel. This way the number of VMs were scaled-out on different nodes,
placing one SQL Server VM on each server node. In this scenario, where there are multiple SQL Server instances, there will be concurrent
checkpoint operations writing to the same disk subsystem. Hence it is required to tune the –k parameter based on the workload. For this test
case, a load of 168,000 transactions per minute was driven on each SQL Server VM, with a total of 672,000 transactions per minute on all four
SQL Server instances. It was determined from the testing that setting the –k parameter to 50 gave a consistent performance on all four SQL
servers.
Technical white paper
Page 11
The graph in figure 7 shows the results of scale-out testing. It shows linear performance gains in terms of batch requests per second on a relative
scale as additional SQL Server VMs were added to service the increasing load. HC 250 for Microsoft CPS Standard system was able to scale-out
easily with the disk subsystem latency well under 10 ms.
% Relative performance
Scale-out
500
450
400
350
300
250
200
150
100
50
0
1 VM
2 VMs
3 VMs
4 VMs
Figure 7. Scale-out performance
Consolidation scenario
This scenario builds on the scale-out scenario from above, with a goal being to simulate consolidation of SQL Server instances running on
multiple underutilized physical servers, onto virtualized machines, thereby reducing data center footprint and sprawl. The testing performed
highlights the platform scalability for ten VMs with different virtual machine configuration and database sizes. The table below lists the number
of VMs and their configuration that were run on the system. SQL Server VMs were distributed across the four nodes of HC 250 for Microsoft CPS
Standard system. Whenever a new VM was added it was always placed on the least loaded server.
Table 3. VM configurations
NO. OF VMS
VM CORE/MEMORY
DATABASE SIZE
6
4 vCPU / 24 GB
55 GB
4
8 vCPU / 48 GB
100 GB
This test case is aimed at driving a varied workload concurrently on multiple SQL Server VMs and tuning them for maximum possible
performance. For this testing, a load of 552,000 transactions per minute was driven across all ten SQL Server VMs. An on average 55,200
transactions per minute on each SQL Server instance. During testing, every time a new SQL Server virtual machine was added, the –k parameter
for each SQL Server already running was tuned to determine an optimum value which resulted in maximum batch requests per second with disk
latencies under the expected limit of 10 ms. From the testing, it was found that a value of –k in the range of 48 to 51 was optimum when all ten
SQL Server VMs were running. The value of the –k parameter for a SQL Server not only depends on the workload on that particular instance, but
also on all other concurrently running SQL Server VMs accessing the same clustered storage subsystem.
Technical white paper
Page 12
Test results showed incremental performance gains whenever a new SQL Server virtual machine was added. The graph below depicts how
performance, in terms of batch requests per second on a relative scale, increased as new SQL Server virtual machines were added.
% Relative performance
1200
1000
800
600
400
200
0
1 VM
2 VMs
3 VMs
4 VMs
5 VMs
6 VMs
7 VMs
8 VMs
9 VMs
10 VMs
Figure 8. Database consolidation
It is important to understand that the total number of SQL Server VMs supported on the HC 250 for Microsoft CPS Standard is entirely
dependent on the workload, database size, and utilization. This testing is just a proof point to demonstrate scaling up to ten SQL Server VMs that
are concurrently serving a reasonable workload mix. It is critical to understand your SQL Server workloads to determine the right sizing and
capacity needs for your HC 250 environment. Fortunately, as you consume the resources on the HC 250, the platform extensibility allows you to
easily add additional nodes to the cluster to grow your resource pools and support more SQL Server workloads.
Analysis and recommendations
Results from the scale-up and scale-out testing demonstrate that small and medium database workloads deployed on HC 250 for Microsoft CPS
Standard system can scale-up as well as scale-out with ease. With HC 250 for Microsoft CPS Standard systems in their data centers, customers
can confidently handle higher loads than usual, by increasing virtual machine compute and storage resources. Also, the HC 250 for Microsoft CPS
Standard system gives an incredible ability to virtualize and consolidate multiple traditional database siloes on just a single system.
If there are many databases with very low utilization, then those databases can be virtualized and put under one SQL Server instance with
enough CPU and memory resources to satisfy its performance requirement. If there are multiple databases with high utilization, then it is best to
put them on different SQL Server instances, which will ensure there are enough resources for optimum performance. It should also be considered
on which HC 250 for Microsoft CPS Standard node the SQL Server instances and other workloads are placed. While creating new virtual
machines for new workloads, place those virtual machines on the least loaded HC 250 for Microsoft CPS Standard node.
When deploying multiple SQL Server instances, it is recommended to distribute them across all four nodes of the HC 250 for Microsoft CPS
Standard system. During testing, it was observed that running multiple instances of SQL Server across different nodes demonstrated better
performance and throughput when compared to running them on a single node.
The Adaptive Optimization (AO) feature in HPE StoreVirtual VSA plays a major role in improving performance of database workloads. Enabling
AO for data volumes that store SQL Server data will ensure optimum performance from the storage subsystem by storing frequently used data
on the SSD tier and lesser used data on spinning media. The SSD tier is limited in capacity and hence other workloads that do not have high
throughput requirements must be stored on separate volumes with AO disabled. This way SQL Server data can make use of AO effectively.
Another important parameter to tune is the –k parameter in SQL Server. This needs to be tuned as per the customer’s environment and
workload. The values specified in this paper are based on the performance obtained during testing and will vary for different SQL Server virtual
machine configurations and for different workloads. For this testing, the HC 250 for Microsoft CPS Standard was used only to run SQL Server
Technical white paper
Page 13
instances, but in customer environments, it may host other application workloads too. In such scenarios, the -k parameter must be tuned based
on that particular environment.
Summary
Customer environments with IT and database sprawl can make use of the HC 250 platform to consolidate and hence reduce both capital and
operational expenditures. Multiple databases can be quickly deployed and managed on the virtual infrastructure powered by the HC 250 for
Microsoft CPS Standard. Testing results have also demonstrated that SQL Server can be scaled-up and scaled-out with ease on the HC 250, to
handle demanding workloads.
The HC 250 system is an excellent platform for virtualization and consolidation. It delivers the right mix of hardware, software and virtualization
technologies that is perfectly suited for small and medium workloads in remote and branch office locations.
Implementing a proof-of-concept
As a matter of best practice for all deployments, HPE recommends implementing a proof-of-concept using a test environment that matches as
closely as possible the planned production environment. In this way, appropriate performance and scalability characterizations can be obtained.
For help with a proof-of-concept, contact an HPE Services representative (hpe.com/us/en/services/consulting.html) or your HPE partner.
Appendix A: Bill of materials
Table 4 lists the major components used for testing.
Note
Part numbers are at time of publication and subject to change. The bill of materials does not include complete support options or other rack and
power requirements. If you have questions regarding ordering, please consult with your HPE Reseller or HPE Sales Representative for more
details. hpe.com/us/en/services/consulting.html
Table 4. Bill of materials
QTY
PART NUMBER
DESCRIPTION
1
N9X97A
HPE HC 250 for Microsoft CPS Standard
4
M0T04B
HPE Hyper Converged 250 Node
4
793028-L21
HPE XL1x0r Gen9 E5-2680v3 FIO Kit
4
793028-B21
HPE XL1x0r Gen9 E5-2680v3 Kit
64
728629-B21
HPE 32GB 2Rx4 PC4-2133P-R Kit
16
781518-B21
HPE 1.2TB 12G SAS 10K 2.5in SC ENT HDD
8
816985-B21
HPE 480GB 6G SATA MU-3 SFF SC SSD
4
665243-B21
HPE Ethernet 10Gb 2P 560FLR-SFP+ Adptr
4
P8Z24A
HPE HC 250 SW for Microsoft CPS Std LTU
1
JG296A
HPE 5920AF-24XG Switch
Technical white paper
Page 14
Resources and additional links
HPE Hyper Converged Systems
hpe.com/info/hyperconverged
HPE Hyper Converged 250 for Microsoft Solution brief
http://h20195.www2.hpe.com/V2/GetDocument.aspx?docname=4AA6-2692ENW
HPE Hyper Converged 250 for Microsoft Data sheet
http://h20195.www2.hpe.com/v2/GetDocument.aspx?docname=4AA6-2656ENW
HPE StoreVirtual VSA
hpe.com/storage/vsa
HPE Hyper Converged 250 for Microsoft Cloud Platform System Standard Manuals
http://h20566.www2.hpe.com/portal/site/hpsc/public/psi/home/?sp4ts.oid=1008873539#manuals
HPE Converged Infrastructure Library
hpe.com/info/convergedinfrastructure
HPE Servers
hpe.com/servers
HPE Storage
hpe.com/storage
HPE Networking
hpe.com/networking
HPE Technology Consulting Services
hpe.com/us/en/services/consulting.html
To help us improve our documents, please provide feedback at hpe.com/contact/feedback.
Sign up for updates
Rate this document
© Copyright 2016 Hewlett Packard Enterprise Development LP. The information contained herein is subject to change without notice.
The only warranties for HPE products and services are set forth in the express warranty statements accompanying such products and
services. Nothing herein should be construed as constituting an additional warranty. HPE shall not be liable for technical or editorial errors
or omissions contained herein.
Microsoft, Windows Server, and Windows are either registered trademarks or trademarks of Microsoft Corporation in the United States
and/or other countries. Intel and Xeon are trademarks of Intel Corporation in the U.S. and other countries. VMware is a registered
trademark or trademark of VMware, Inc. in the United States and/or other jurisdictions.
4AA6-5301ENW, April 2016
Download PDF
Similar pages