Your Virtualization Community: Optimize the performance of your virtual environment.

Performance Improvement Due to ESX 3.5?

  • I recently got an email from asking about the testing results published on this page. The question was how much of the improvement in peformance could be linked to the upgrade from ESX 3.0.1 to ESX 3.5?

    The answer is that I didn't do the R900 test with ESX 3.0.1, so it is not possible to say how much of the performance advantage that the R900 has is due to ESX 3.5. I can say that with the upgrade from ESX 2.5 to ESX 3.0.1 I did some testing with the same workload on the same server and found that performance was about the same. I would guess that most of the performance advantage is due to the R900 based on this previous experience.

  • When the R900 VMmark score was first released November 19th, 2007 it ran 12.23 @ 8 tiles ( We ran it again and published the score on July 8th, 2008 and got 14.05 @ 10 tiles (

    The latest test ran on ESX 3.5 update 1 (the first ran on ESX 3.5). Minor differences on a couple other items, but for the most part the big difference was Update 1 (as far as I can see).

    VMmark Scores -
  • This is obviously just informational, not necessarily a direct reply. I would love to know exactly what made the big jump (if it was more configuration or more Update 1).
  • The big difference in configuration appears to be the storage. On the first test a Dell|EMC CX300 with RAID 10 LUNs was used. On the better performing second submission a Dell|EMC CX3-40 with RAID 0 LUNs was used. I'm not sure if this what actually resulted in the improved performance. I'm going to forward this thread to one of the guys here that was involved with the testing.

  • Well, I talked to one of our lead guys in the performance testing lab and he thinks that a big part of the gain was due to ESX 3.5 Update 1. The problem is that there were two factors that changed between the two tests - the SAN and ESX 3.5 Update 1. So without some additional testing that would isolate these changes it is really not possible to say which is more responsible for the performance increase. I would bet that the gain is due to both, but I can't say which had a bigger impact.

  • Both sounds reasonable...

    I would love to see it with EqualLogic.
  • We just got some more EqualLogic arrays in the lab. And I might be working on a new virtualization performance test using those for the storage. I'll post more to the Virtualization Frontier blog when I am a bit further along.