Clint Huffman Microsoft Premier Field Engineer (PFE) Microsoft Corporation SESSION CODE: VIR315 Kenon Owens Technical Product Manager Microsoft Corporation Windows Server 2008 R2 Hyper-V Performance Analysis: How You Can Get the Most out of Hyper-V
Hyper-V can’t run that many VMs per box (I don’t get the density) Hyper-V can’t perform under load I can’t/won’t run my business critical applications like SQL, SharePoint and Exchange on Hyper-V Try Hyper-V R2, it makes a big difference. Don’t compare against R1 of Hyper-V Windows Server 2008 R2 has incredible performance Hyper-V R2 is a phenomenal platform for your Microsoft business critical apps and your apps from other ISVs as well
iSCSI Performance with Intel® G NIC with VMDq, Intel® Xeon 5580 Platform, Windows Server 2008 R2 and R2 Hyper V Performance tests and ratings are measured using specific computer systems and/or components and reflect the approximate performance of Microsoft Intel products as measured by those tests. Any difference in system hardware or software design or configuration may affect actual performance. Buyers should consult other sources of information to evaluate the performance of systems or components they are considering purchasing. Read/Write IOPs and Throughput Test
) SQL Server Performance: SLAT impact Results: Increased throughput with consolidation Near linear scale in throughput with no CPU over-commit Improved performance with Windows Server 2008 R2 and SLAT processor architecture % CPU Throughput (Batch requests/sec) Configuration: OS: Microsoft® Windows Server® 2008 R2 Hyper-V™ Hardware: HP DL585 (16 core) with SLAT (Second Level Address Translation) processor architecture, HP EVA 8000 storage Virtual Machines: 4 virtual processors and 7 GB RAM per virtual machine; Fixed size VHD Relative Throughput for Windows Server 2008 Heavy Load Moderate Load Low Load CPU over-commit Almost Linear Scale No CPU over-commit Almost Linear Scale No CPU over-commit Batch requests/sec/%CPU Windows Server 2008 R2
* Used pre-release software (ESX Build ) and the iSCSI initiator in the VM (not pass-through). For Hyper-V, they used pass-through drives for data. **HP would likely see similar results when replacing their own older systems with new HP machines ***Consolidated and Virtualization Hardware platforms were the same in both tests Given the minor differences, Hyper-V is an extremely viable alternative to ESX, capable of competitive levels of VM density, even for customers who already have sunk costs in ESX licenses. For customers embarking on new virtualization deployments, Hyper-V is VERY competitive! For More Information Check out These Papaers: L385onto5PEM610.pdf L385onto5PEM610-vSphere.pdf
vSphere 4Hyper-V R2 5 Dell Servers (35 VMs)13,798 OPM15,734 OPM Power (Loaded)4,503 Watts4,265 Watts Power (Idle) 3,611 Watts3,472 Watts
Windows Server 2008 R2 Hyper-V Performance Analysis: How You Can Get the Most out of Hyper-V Clint Huffman Microsoft Premier Field Engineer (PFE) Microsoft Corporation SESSION CODE: VIR315
Root Partition I/O Stack Drivers Child Partition I/O Stack VSCs Server Child Partition I/O Stack VSCs Server Hypervisor DevicesProcessorsMemory VMBus Shared Memory VSPs OS Kernel Enlightenments (WS08+)
DISK MEM NET CPU
DISK MEM NET
X
Hyper-V Processor Utilization % Processor Time performance counters on the root partition are not accurate. Use the % Guest Run Time and % Total Run Time counters Root Partition (physical host) \Hyper-V Hypervisor Logical Processor(*)\% Total Run Time 95% Virtual BizTalk % Processor Time 100% Root Partition % Processor Time Virtual BizTalk % Processor Time 100% \Hyper-V Hypervisor Logical Processor(*)\% Guest Run Time 90% % 5%4%3%
BizTalk Server Virtual Processors 0 0 BizTalk Server Virtual Processors 0 0 BizTalk Server Virtual Processors 0 0 BizTalk Server Virtual Processors 0 0 Logical Processors to 1 Ratio2 to 1 Ratio BizTalk Server Virtual Processors BizTalk Server Virtual Processors BizTalk Server Virtual Processors BizTalk Server Virtual Processors Logical Processors Over utilized? Under utilized?
Runs even better on R2!
Not optimal Optimal Runs even better on R2!
\Processor(*)\% Processor TimeMinAvgMax Std Deviation 10% of Outliers Removed Physical SQL Hyper-V Hypervisor Logical Processor(_Total)\% Total Run Time Virtual SQL
XX
Optimal
DHCP DNS IIS Directory Services … X
Higher Is Better Lower Is Better SSD has more throughput! SSD has faster response times Runs even better on R2!
Higher is better! Runs even better on R2!
Root Partition (physical host) Root Partition (physical host) – 16GBs of RAM Virtual Guest RAM: 2GBs Committed Memory Pagefile.sys Committed Memory Virtual Guest RAM: 4GBs Committed Memory Pagefile.sys
Root Partition (physical host) Root Partition (physical host) – 16GBs of RAM Physical Computer RAM: 2GBs Committed Memory – 2GBs Pagefile.sys Committed Memory – 2GBs Virtual Guest RAM: 4GBs Committed Memory Pagefile.sys
Not optimal Optimal
Runs even better on R2!
Sign up for Tech·Ed 2011 and save $500 starting June 8 – June 31 st You can also register at the North America 2011 kiosk located at registration Join us in Atlanta next year