This Storage Intelligence (StorInt™) dispatch covers recent Microsoft Exchange 2010 Solution Review Program (ESRP) v3.0 results. Since we last discussed this 1K-to-5K mailbox category nine months ago, there have been at least four new ESRP submissions and another eight in other categories. Future dispatches will report on the over-5K and 1K-and-under mailbox categories, but all previous ESRP V2 and ESRP V3 performance dispatches are available on SCI’s website.
Latest ESRP V3.0 results
We start our ESRP analysis with Exchange database transfers/second/spindle.
Figure 1 Top ESRP database transfers per second per disk spindle
For Figure 1, two of the Dell results are new to this analysis. Specifically, the new #1 from Dell EqualLogic PS with 5000 mailboxes and the #4 Dell PowerVault MD1220 with 5000 mailboxes using 10Krpm SAS. Often, speed of disks used in storage can have a significant impact on standings here but the middle four (#4-7) and last one (#10) used either 10Krpm disks (#4-5) or 7200rpm disks (#6-7 & 10) and all the rest were 15Krpm. So disk speed is not the only factor influencing these results.
We now turn to average Log Playback Time results.
Figure 2 Top ESRP Log playback time
In Figure 2 lower is better. The #7, 8 & 9 rankings on this chart are all new to this analysis. We have always been big fans of log playback time because ESRP performance is so configuration driven. Log playback is just 1MB of log data that is played back against the database as fast as the storage system can handle it. ESRP parameters don’t seem to have much impact on this performance
The interesting thing about Log playback results is that there’s another nice mix of 4-iSCSI (#1 & #7-9), 3-SAS (#3 & 5-6) and 3-FC (#2, 4 & 10) attached storage systems. It doesn’t seem to make much difference, at least at this level, what storage protocol you use, as they all can perform well.
We now show (un-normalized) database transfers per second.
Figure 3 Top ESRP database transfers per second (un-normalized)
Figure 3 shows the Top 10 database transfer performance results. New results include #6 for Dell EqualLogic with 15Krpm and #8 for Dell PowerVault MD1220 iSCSI attached storage with 10Krpm SAS drives. Not much to discuss here other than the amazing performance of the #1 HP P2000 SAS attached storage which no one else (including HP P2000 iSCSI & FC systems) seems able to even do half as well. Must be some secret here, lets see all 3 P2000 used the same drive (15Krpm, 300GB drives) and almost the same number ( 94, 94 & 90 respectively), but #1 obtained almost 3 times the database transfers per second of the #2 iSCSI config and 4.7X the FC configuration?
Next, we return to another of our perennial favorites, database access latencies.
Figure 4 Top 10 Database read access latency
In Figure 4 lower is better. This chart is sorted by read access latency but we show write and log latency as well. The only new result on this chart is HP P4500 (Lefthand) storage coming in at #10 with a read latency at 16 msec. just over double the #1 Fujitsu read latency of 7.7msec. I would say that one reason for their #10 read latency showing was the 7200rpm drives but for the fact that the #1 Fujitsu storage also used 7200rpm drives. The surprise here is that except for #1, the next 4 ranked systems (#2-5) all used 15Krpm systems, then the next 2 (#6-7) used 10Krpm drives, and then the final 3 all used 7200 rpm drive.
Finally, we introduce a new chart used in to compare storage interface performance.
Figure 5 Scatter plot: ESRP KB of log processed per second
We have plotted all results for this category on this scatter plot to attempt compare how well different storage interfaces can. A couple of caveats first, FC had only 4 results and iSCSI had only 7 so each of these are probably not statistically relevant. Even SAS only had 11 results so there may not be much to show there either. Nonetheless, it would seem that SAS is performing better than iSCSI at least in KB log processed per second, and FC is doing better than both SAS and iSCSI.
Reiterating a theory from our last report, there are more mysteries in ESRP results than in any other performance benchmark we look at. Not many of the rankings would seem to fit the pattern that faster disks means better performance. Of course, the fact that there are so many open questions on ESRP performance results just makes them more interesting, at least to us.
The 1K-to-5K mailbox category hasn’t been all that active recently with just four new submissions out of a total 22. But there does seem to be some modest interest here at least as compared to the 1K-and-under category.
Finally, as discussed in prior dispatches, ESRP/Jetstress results seem designed to be difficult to compare but in our view, merit the effort. As such, feel free to contact us with any constructive ideas on how to improve. In that regard, our contact information can be found below or on our website at SilvertonConsulting.com.
[This performance dispatch was originally sent out to our newsletter subscribers in April of 2012. If you would like to receive this information via email please consider signing up for our free monthly newsletter (see subscription request, above right) and we will send our current issue along with download instructions for this and other reports. Also, if you need an even more, in-depth analysis of SAN storage system features and performance please see our SAN Storage Buying Guide available on our website.]
Silverton Consulting, Inc. is a Storage, Strategy & Systems consulting services company, based in the USA offering products and services to the data storage community.
 ESRP results from http://technet.microsoft.com/en-us/exchange/ff182054.aspx, as of 24 April 2012