vSphere Beats Hyper-V in Side-by-Side Test

Yesterday, I was called by a PR guy I know who works with VMware's PR agency. He was returning my call to him in which I suggested having VMware write a piece addressing a very unfavorable white paper about VMware's vSphere pricing that Microsoft was circulating around the time of VMworld. The PR guy told me that it was not VMware's policy to respond to the attacks Microsoft is frequently making.

Sort of a stay-above-the-fray type of attitude. It made sense to me because one, there are so many people saying nasty things about VMware all the time that the company would have to hire a dedicated damage-control team just to answer all the naysayers--that's the way it goes when you dominate a market--and two, VMware is keeping an eye of Microsoft, but Paul Maritz is not losing any sleep over his former employer.

My PR guy did, however, refer me to VMware's Virtual Reality blog, which he said is sometimes used as a forum for competitive jousting. When I went to it today, I found an August 26 posting from Eric Horschman headlined "It's no surprise that vSphere 5 holds up under the pressure, but what about Hyper-V?" Horschman basically tees up a white paper entitled, "VMware vSphere outperformed Microsoft Hyper-V R2 SP1 by 18.8%," in order to see how Microsoft's Dynamic Memory would hold up against vSphere in a side-by-side VM density test.

VMware hired Principled Technologies to do the side-by-side comparo of vSphere 5 and Hyper-V R2 SP1 throughput when running a SQL Server workload under high VM densities. Said Horschman, "They used the well-respected DVD Store Version 2 benchmark to measure total throughput delivered by a host running 24 VMs, and then 30 VMs. With 24 4GB VMs, the 96GB host was just reaching full memory commit, and 30 VMs pushed it to 25% memory overcommit--familiar territory for vSphere users."

In a nutshell, here's what Principled Technologies reported: "We tested two hypervisors--VMware vSphere 5, and Microsoft Hyper-V, part of Windows Server 2008 R2 SP1--to compare their performance and ability to manage resources at high levels of utilization. When running 30 virtual machines each, VMware vSphere 5 outperformed Microsoft Hyper-V SP1 by 18.9%. VMware vSphere 5 also allocated server resources among individual VMs more fairly than Microsoft Hyper-V R2 SP1 (39.2 percent less variation), resulting in more consistent performance across all virtual machines. Finally, VMware vSphere 5 did a better job of scaling from 24 to 30 VMs with our workload: Overall server performance increased by 11.2 percent with VMware vSphere 5, whereas performance decreased 3.3 percent with Microsoft Hyper-V R2 SP1."

There's a lot more good information in this report, which you can download from the bottom of Horschman's blog. And, oh by the way, the next blog down is entitled "Setting Microsoft Straight on the VMware Service Provider Program (VSPP).

So much for staying above the fray.

Posted by Bruce Hoard on 09/15/2011 at 4:05 PM