Home > SVC/V7000U > IBM V7000 sets the bar with SPC1 all Flash results

IBM V7000 sets the bar with SPC1 all Flash results

Last week IBM published some very interesting results on the Storage Performance Council website.  Using the SPC-1 test method, IBM raised more than just a few eye brows.  IBM configured 18 200GB SSDs in a mirrored configuration which was able to attain 120k IOPS with less than 5 ms response time even at the 100% load.

IBM used an all SSD array that fit into a single 2U space and mirrored the drives in RAID 1 fashion. These were all put into a pool and 8 volumes were carved out for the AIX server. The previous SPC1 run IBM performed used spinning media and virtualized the systems using the SAN Volume Controller virtualization engine. This gave IBM the top spot in the SPC1 benchmark with over 520,000 IOPS costing a wopping $6.92 per IOP.

This has been compared to the Oracle/Sun ZFS 7420 who published 137k IOPS late last year.  When matched to the IBM V7000 results we see the V7000 was $181,029  with roughly $1.50 per IOP compared to the SUN 7420 who came in at $409,933 and $2.99 per IOP. The V7000 was able to perform 88% of the work at 44% of the price. Oracle has not come back with any type of statement but I can only bet they are seething over this and will be in the labs trying to find a way to lower their cost while still performing.

The SPC1 benchmark tests the performance of a storage array doing mainly random I/O in a business environment. It has been called out for not being a very realistic workload as it tends to carter to higher end, cache heavy systems.  Now we have a mid range box that not only has the most IOPs trophy but also wins in the ‘Best Bang for your Buck’ category too.

I would have liked to see what the results would have been with the addition of compression to the software feature stack. This  going to be a game changer for IBM as the inline compression of block data is way beyond what other vendors are doing.  Couple that with the virtualization engine and now I can compress data on anyone’s storage array. The V7000 is definitely becoming a smarter storage solution for a growing storage market.

 

Advertisements
  1. data_integrity
    June 27, 2012 at 5:03 am

    The comparison with V7000 and the 7420 is just false. The V7000 with only SSDs should maybe be compaired to Oracles flasharrays. The 7420 will outperform the V7000 anyday. Sorry for your lack of knowledge. The V7000 holds barely 1TB of storage while the 7420 holds 23TB. And secondly why would you put your data on a storage array without end to end data integrity?
    So this is just nonsense.

  2. data_integrity
    June 27, 2012 at 5:12 am

    Oh and secondly compression, dedup and all the other goodies are included in the 7420. And the Gui in V7000 is so ugly compaired to 7420. Anyone who gets their head out of the ground will realize that their is nothing new with IBM storage. Even Netgear has realized the power of ZFS and started to create their own ZFS appliance.
    So my what I’m trying to say is that IBM storage is loosing ground and so is HP. These two companies atleast in the storage segment try to profit on onld technology and it makes me mad.

  3. September 16, 2012 at 5:32 pm

    I carry an Oracle batch, just to let you know before you read the rest of my comment. Any SSD only system can easily outperform a system composed of regular disk drives on price/iops metric.

    If I were you I’d try to anwser the following questions :

    1) Why is V7000 beaten on SPC-1 benchmarks that uses regular spindles vs 7420 by a so wide margin, despite the fact that it uses far cheaper components?

    2) How can you get only 6.6k IOPS per drive on the SSD only system? V7000 is almost wasting %70 of the drive performance. What is limiting the system, fw, cpu power, backend, all of them?

    3) How much usable capacity does 18x 200GB disks give to your customer? and if it’s OK for your customer, how do you think a 7420 with 2TB cache will perform in this environment?

    best regards
    Mertol

    • October 8, 2012 at 9:28 am

      1. not sure i get your point other than we beat oracle on performance and price. The SPC-1 benchmark is just that, a water mark to hold everyone to the same level of scrutiny.
      2. RAID and spares?
      3. not sure, does oracle have a SPC1 out for it that we can compare to?

      Rich

      • October 8, 2012 at 2:30 pm

        @richswain
        Just to make things clear. If we imagine that a customer with a Oracle 7420 can hold their working set in ARC (RAM) or L2ARC (SSD) which is what you usually calculate with when going with a ZFS appliance then your performance is unbelievable. Imagine boot times for virtual machine when having hundreds of gigabytes of RAM cache. They boot extremely fast.

      • October 9, 2012 at 4:29 pm

        Hi Rich ;

        Take an other look at the SPC site, you will notice an other benchmark of V7000 with spindles. It’s doing 2.5x less IOPS per USD than a 7420, and with just %16 more drives it achieved 2.6x more performance. And if you compare the latency graphs you will notice that 7420 offered far better latency though out the range. good luck explaining this…

      • October 10, 2012 at 6:34 am

        yes, the V7k SPC with spinning disk was less. BUT the idea is the V7k can achieve more with a full SSD based system than that of the 7420. My problem with the 7420 is the idea that you have to use DRAM or flash cards to get the performance that we can get with eMLC SSD drives. I would also point out the fact that you have buy almost 3X the footprint to get that performance vs the V7k. The other part is do we really compare the two? or is that all Oracle really has to offer? 😉

      • October 10, 2012 at 1:26 pm

        Rich; if you really want to compare all SSD systems, try comparing the V7000 with the Oracle F5100 than, which offer not only 1/4 $ IOPS but also 2X + performance. It’s too meaningless to compare a V7000 with very little usable capacity to a 7420 with a lot of usable capacity. As I stated before V7000 achieve only 6.6k IOPS per SSD drive, I see nothing to brag about here.

  4. October 10, 2012 at 11:12 am

    @richswain
    Yes but you don’t have to fill the 7420 with these devices to achieve great performance. And I would actually say its sad that the V7000 is using eMLC which with intesive writes will wear out. With a 7420 or other ZFS appliance you will use SLC for write cache and eMLC for secondary cache (L2ARC). The V7000 can never be justified against a 7420 or any other ZFS appliance.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: