This Storage Intelligence (StorInt™) dispatch covers Storage Performance Council (SPC) results. There have been three new SPC-1 and no new SPC-2 submissions since our last report in May. The new SPC-1 results are from HDS Hitachi Unified Storage (HUS) VM, Fujitsu DX600 S3 and Dell EqualLogic PS6210XV. Two of our top ten charts have changed and we present them below.
We begin our discussion with LRT™ (least response time) top ten resultsshown in Figure 1.
First, all of our top ten LRT systems have a majority of their ASU capacity in SSDs or Flash storage. The new submissions from HUS VM (#4) and Fujitsu DX600 S3 had all of their storage in Flash modules or SSD. In fact, the HUS VM used 1.6TB flash modules, a Hitachi design and the Fujitsu DX600 S3 had 400GB SSDs.
It’s sad to say but now, there aren’t any hybrid storage (with disk and less than 50% of ASU capacity in flash) systems on our top 10 LRT results. It’s not clear why a properly designed hybrid system couldn’t do well here. But even a few misses would skew average response time up and recall that LRT is measured at 10% IOPs load.
The purpose-built, TMS (now IBM FlashSystem storage) all flash storage arrays continue as the top 3 systems to beat in LRT. But the HUS VM with its Hitachis designed, FlashModule storage is closing in and the surprisingly quick, Fujitsu DX600 S3 using “standard” SSD storage is not far (~0.01msec) behind.
Next, we turn to IOPS™/drive top ten chart in Figure 2.
In Figure 2 we see the Dell EqualLogic PS6210XV came in at tenth place, with a little under 350 IOPS/drive. Another interesting item is that the Dell storage was an iSCSI array. We seldom see iSCSI storage in SPC-1 benchmarks. I do believe this is the only one I have come across, but there may be others that are in inactive or yet to be published (still in audit cycle) status.
In our SAN-NAS Buying Guide we publish a chart that compares iSCSI, FC and SAS IO activity and typically find that FC does better than iSCSI and iSCSI better than SAS in IOs/drive activity that we report on. Unclear why Dell EqualLogic did so well here but most disk drives are incapable of sustaining over 200 IO/second, so I can only assume their good showing depends on more effective caching and better parallelism throughout their storage system. Now how about releasing a SPC-2 benchmark on an iSCSI storage system.
It’s great to see any new SPC-1 disk-only results and iSCSI results even better. The fact that the LRT top ten doesn’t even show a hybrid storage system anymore probably says more about the niche that all flash arrays are used for these days (consistently low responsiveness) than anything else. In our OLTP ChampionsCharts™ we have separated all or majority flash arrays into their own chart because they were skewing comparative response time and IO operation per second results.
As always, suggestions on how to improve any of our performance analyses are welcomed. Additionally, if you are interested in more block performance details, we now provide a fuller version (Top 30 results) of some of these charts and a set of new OLTP, Throupgput and Email ChampionsCharts™ for enterprise, mid-range and SMB SAN storage, in our recently updated (February 2019), SAN Storage Buying Guide available from our website.
[This performance dispatch was originally sent out to our newsletter subscribers in August of 2014. 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 Storage, Strategy & Systems consulting services company, based in the USA offering products and services to the data storage community.