This Storage Intelligence (StorInt™) dispatch covers Microsoft Exchange 2010 (E2010) and Exchange 2013 (E2013) Solution Review Program (ESRP) results. Since we last reviewed this 1001 to 5000 mailbox category nine months ago, there have been only two new entries, the HP D3600 disk enclosure/P441 RAID controller and the Dell PowerEdge R730xd storage. Both these configurations managed to perform well enough to crack into our top ten best performing ESRP backup systems.
Latest ESRP V3.0 & 4.0 (E2010&E2013) performance
In Figure 1 we show an ESRP reported, top ten MB/sec./server backup performance metric.
As can be seen in Figure 1, both the new Dell PowerEdge R730xd (#2) and HP D3600 disk enclosure P441 storage server (#10) did well on ESRP backup data transfer performance per server.
The HP D3600/P441 reached ~1.1GB/sec in backup throughput using a single D3600 drive shelf attached to a P441 RAID controller, with 11 6TB, 7200 RPM disk drives and 4GB of “flash backed write cache” (FBWC, in the P441 controller). This RAID controller card was attached to HP ProLiant DL360 Gen 9 Exchange server(s). HP’s (LSI’s) FBWC is used for both write back and read-ahead caching.
Dell, with their 2.3GB/sec backup per server, doesn’t list its configuration as a drive enclosure, but essentially had the same setup as the HP solution. The Dell PowerEdge R730xd Exchange server(s) incorporated a PowerEdge H370P Mini RAID controller card to attach a drive shelf with 14 4TB, 7200 RPM disk drives. The Dell H370P only had 2GB of “flash backed cache”.
Not clear why the Dell JBOD configuration did much better than the HP configuration but the higher capacity drives (6TB vs. 4TB) and the lower drive counts (11 vs. 14) probably hurt Its backup performance. Although, the additional FBWC capacity in the HP solution, with proper smarts, should have helped overcome this deficit.
In general, we don’t like the MB/sec/server metric because it under estimates storage backup throughput when more than one Exchange server is used. But, at least for these two solutions, with only one (active) server operating during the benchmark, this was not an issue. In this case, the MB/sec/server provided a correct estimate of these storage solution’s ESRP database backup performance.
In Figure 2 we show a SCI derived metric, top ten ESRP total database backup throughput.
For Figure 2, we sum up all the database backup activity for a solution to compute a total backup performance metric. Here, one can see how well the new Dell R730xd solution (#6 above) did against the other systems with superior backup performance. Again most of these solutions were Exchange servers using SAS connected JBOD, with three exceptions being the HDS HUS 130, HDS HUS 110 and Fujitsu ETERNUS DX200 S3 (#4, #5 & #10 respectively).
However, the Dell solution probably suffered from low drive counts (the Fujitsu also had 14) and the largest capacity drives (4TB vs. 1 to 3TB for all the others). What’s surprising is even with those two disadvantages, it managed to provide middle of the pack backup performance. So they must be doing something right.
Unfortunately, none of our other top ten ESRP performance charts have changed for this 1001 to 5000 mailbox category. So if you are interested in seeing any of these charts, we suggest you refer to prior ESRP reports located on the Dispatches page on our web site.
There hasn’t been as much new activity in this 1001 to 5000 mailbox category over the past 9 months as we would like to see. But what we have seems to be doing well in backup performance, even with jumbo sized drives. It’s good to know that current large capacity drives are at least providing comparable sequential performance (for backup throughput) as earlier drives.
On the other hand, it appears that RAID controller card flash caching does nothing for the other ESRP IO performance activity such as database (random) transfers, log (sequential & random) playback, or database (read & write) and log (write) response times. All these other activities have significant amounts of random IO, which current controller card flash caching doesn’t seem to help. As a result, large drives supply poor Exchange IO performance (except for backups), at least in JBOD/RAID controller attached storage configurations.
Constructive comments on how to improve our analyses for Exchange or any of our performance reports are always welcome. Moreover, if you detect errors in this or any of our other performance reports, please do let us know and we will correct it as soon as possible.
If you received this report from someone else, please consider signing up for your own free copy of SCI’s StorInt™ monthly newsletter using the QR code below right. We provide commentary on recent major storage system announcements as well as analyze current storage system performance results/benchmarks like this each month. We plan to analyze Microsoft ESRP results again in three months.
[Also we offer more block storage performance information plus our OLTP, Email and Throughput ChampionsCharts™ in our recently updated (February 2019) SAN Storage Buying Guide, or for more information on some select ESRP performance results please see our recently updated (December 2019) SAN-NAS Storage Buying Guide, both of which are available for purchase on our website.]
[This performance dispatch was originally sent out to our newsletter subscribers in October of 2015. 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. Dispatches are posted to our website at least a quarter or more after they are sent to our subscribers, so if you are interested in current results please consider signing up for our newsletter.]
Silverton Consulting, Inc., is a U.S.-based Storage, Strategy & Systems consulting firm offering products and services to the data storage community.
 ESRP results from https://technet.microsoft.com/en-us/office/dn756396.aspx as of 28Oct2015