VMware Releases Storage Protocol Performance White Paper
February 7th, 2008 |
Hot off the press. This document compares the performance of four storage options for ESX 3.5; FC, Hardware iSCSI, Software iSCSI, and NFS. I haven’t had a chance to thoroughly digest it yet, but I look forward to going over it in more detail this weekend.
You can get the white paper here.
11:24 am
[...] Thomas, it looks like VMware has published a storage performance white paper comparing Fibre Channel, hardware iSCSI, software iSCSI, and NFS. I just finished reviewing [...]
8:18 pm
[...] has just released a paper entitled Comparison of Storage Protocol Performance (seen at Scale the Mind and blog.scottlowe.org); maybe this will help deflate some of the too-often repeated speculation [...]
6:03 am
Interesting that they published, but not very surprising. FC is faster but costs more. I was more interested in how closely the hardware and software iSCSI compared with each other, FC and with NFS.
3:47 pm
Yes, not very surprising based on the scenario they tested. There are some other elements that I would like to see tested (more on this in an upcoming post).
8:40 pm
[...] of some recent storage tests over on their blog today. This comes just a few days after VMware released a white paper which compares the performance of different storage protocols in ESX 3.5. The information posted today details the scaling performance of VMFS. The tests were run with FC [...]
6:40 pm
[...] regarding storage performance with ESX 3.5. The first bit of information was in the form of a white paper which compares storage protocols in ESX 3.5. The second, was a blog post over on the VMware performance team blog which provides the results of [...]
1:59 pm
The whitepaper published leaves a lot to be desired and has several flaws:
1) As everybody has already mentioned doing 100% I/O of a workload (reads/writes/random/seq) is not representative of any real environment
2) Using a single ESX host is not an implementation customers typically deploy. How many customers with a single ESX server do you know?
3) Using RDM only lends itself to speculation as to why VMFS was absent from the tests given that most VMware customers deploy VMFS
If you gonna put out a paper comparing protocols, then it ought to be done correctly and in a manner representative of a typical VMWare customer deployment. Also, since ESX enables scale out clustering, and VMFS is the primary mechanism in doing so, a good test would be to compare the scaling aspect of it and what does performance look like in that scenario