Home > Blogs > VMware VROOM! Blog > Monthly Archives: March 2012

Monthly Archives: March 2012

NFS Performance on VROOM! Videos

In this episode of VROOM! Videos we discuss NFS performance on vSphere.  We cover the basics of NFS, some of the key considerations when testing NFS performance, and finally a summary of the type of results we have seen recently.


This is episode 4 of season 1 of VROOM! Videos.  They premier on VMware Community TV on Fridays with a live chat and live video intro.  All episodes are available here on VROOM! and on VMware TV.

VROOM! Videos Talks About Hadoop and Big Data

In this episode of VROOM! Videos we talk about big data and Hadoop on vSphere.  More details on the great performance work that Jeff has done with Hadoop can be found in an earlier VROOM! blog post.


New episodes of VROOM! Videos are premiered live on VMware Community TV on Fridays (1 pm Eastern).

VMware View 5 outperforms Citrix XenDesktop

Principled Technologies has just released an interesting whitepaper comparing the performance of VMware View 5 with Citrix XenDesktop 5.5 — the doc can be found here. In this paper, View and XenDesktop are compared using LoginVSI, for both the light and medium workloads, and both network and CPU utilization data are presented.

The key benefit of this study is it compares the two VDI technologies in an apples-2-apples manner. Out of the box, View and XenDesktop deliver different levels of desktop image quality — with XenDesktop delivering a lower image quality by default. Accordingly, for a meaningful comparison between VDI technologies, it is critical to adjust the quality settings to ensure that both VDI solutions are delivering comparable quality. In this paper, Principled Technologies, provide data for both high quality imaging (increasing XenDesktop image quality to match View's default) and lossy quality (decreasing View image quality to match XenDesktop) benchmark runs. In both these usage scenarios, View is found to achieve lower resource utilization in the majority of runs! The only exception is runs with in which XenDesktop flash redirection is enabled — Principled Technologies have not configured View video as discussed in our prior blog postings (which, in many instances, is capable of bringing the bandwidth associated with server-rendered flash content close to redirected streams).

Accordingly, the main take-away from this study is that it is important for customers to determine what desktop imaging quality their particular usage scenario demands and to configure their VDI solutions appropriately, both in any bake-off performance testing and in their final VDI deployment. 

Finally, it is important to note that this study using 3rd party workload simulation tools, closely matches the findings that we presented at VMworld 2011 using VMware View Planner.

A Conversation About 1 Million IOPS

The second episode of VROOM! Videos is a discussion about the testing that we did to achieve 1 million IOPS on a single vSphere5 host.  


For additional information about the 1 Million IOPS test please read the earlier VROOM! blog post.  New episodes of VROOM Videos are premiered on VMware Community TV on Fridays at 10 am Pacific.

VROOM! Video Episode 1 – VMmark 2

The performance team at VMware is starting a new series of videos called VROOM! Videos. Each video will focus on a single topic and be five to ten minutes in length, to keep things focused. The videos are done by the engineers in informal settings and are meant to be fun discussions. They are the type of conversation you might have at lunch if you asked somebody what they have been working on lately. The first episode is a conversation about VMmark 2.

VMmark 2 website

We are also doing a live launch for each of these videos on Fridays on VWware Community TV. (1pm EST, 10am PST) The premiere for each video will include a live introduction to the video and a live chat with the engineers.