IBM FlashSystem

IBM FlashSystem

Find answers and share expertise on IBM FlashSystem


#Storage
 View Only

Good News and Bad News

By Tony Pearson posted Wed January 30, 2008 01:52 PM

  

Originally posted by: TonyPearson


Last July, IBM and EMC traded blog postings over SPC-1 benchmark results. Fellow EMC bloggerChuck Hollis wrote his post [Does Anyone Take The SPC Seriously?]. Here is an excerpt:
I think most storage users have figured this out. We've never done an SPC test, and probably will never do one. Anyone is free, however, to download the SPC code, lash it up to their CLARiiON, and have at it.

I responded with [Getting Under EMC Skin], and then followed up with a series explaining IBM SVC and SPC benchmarks here:

So what is the good news?Yesterday, our friends at NetApp took up Chuck's challenge and posted results on their FAS3040 as well as their EMC CLARiiON devices. IBM sells the FAS3040 under the name IBM System Storage N5300 disk system. Knowing that NetApp maintains excellent performance when it is doing point-in-time copies, NetApp ran both with and without on both boxes. I include DS4700 and DS4800 as well for comparison purposes, but only have them without FlashCopy running.

ProductModeSPC-1 IOPS
IBM DS4800No FlashCopy45,014
NetApp FAS3040 (IBM N5300)No SnapShot30,985
NetApp FAS3040 (IBM N5300)With SnapShot29,958
EMC CLARiiON CX3-40No SnapDrive24,997
IBM DS4700 ExpressNo FlashCopy17,195
EMC CLARiiON CX3-40With SnapDrive8,997

One would expect some performance degradation with a box running point-in-time copies at the same time it is reading and writing data, but NetApp/IBM N5300 does not degrade by much, but EMC's drops a significant amount.

So what is the bad news? Last October, I welcomed HDS USP-V to the [Super High-End Club], but now we need to invite Texas Memory Systems as well.In 2006, I posted [Hybrid, Solid State and the future of RAID], and poked fun at Texas Memory Systems using the slogan "World's Fastest Storage", which at the time that honor belonged to IBM SAN Volume Controller instead.The VP of Texas Memory Systems, Woody Hutsell, explained the only reason their solid-state disk system, RAMSAN-320, didn't have faster results is that they didn't have the fastest IBM server to run against it. It may not surprise you that nearly everyone's SPC benchmarks use IBM servers because IBM has the fastest servers as well. I didn't have a million-dollar System p UNIX server to send Woody for this, but it looks like they have finally gotten one, and a new RAMSAN-400 device, as they have posted their latest results.

ProductModeSPC-1 IOPS
Texas Memory Systems RAMSAN-400Cache only291,208
IBM SAN Volume Controller 4.2Cache/External Disk272,505
HDS USP-VCache/Internal Disk200,245

EMC doesn't publish numbers for their Symmetrix box, despite their announcement of faster SSD drives. They claim that SSD drives make their overall disk system performance faster, but without SPC benchmarks, we will never know. If you have a Symmetrix, this YouTube video may help you decide where it belongs:

YouTube Video: Recycling the Symmetrix



You can read all the[SPC-1 Benchmark Results]on the Storage Performance Council (SPC) website.

technorati tags: , , , , , , , , , , , , , , , , , , , , , , , ,

5 comments
12 views

Permalink

https://community.ibm.com/community/user/blogs/tony-pearson1/2008/01/30/good-news-and-bad-news

Comments

Archive User
Thu February 07, 2008 01:04 PM

BarryB,I already corrected you on your original post, but just in case someone doesn't link through and read all the commments: the DS6000 is not dead. It is alive and well, we have sold THOUSANDS of these systems.
As for disposal of old disk systems, I believe there are a lot of disk systems, from all vendors, that were purchased in anticipation of the Y2K scare, and those are now being disposed of. Like EMC, IBM also has services to make sure that batteries or other toxic chemicals do not pollute the environment. This could explain partially why storage is doing so well despite the rest of the economy doing poorly.
-- Tony

Archive User
Sun February 03, 2008 06:20 PM

Netapps is just showing off their flashcopy algorithm: instead of copy on write like the rest of the world uses, they use a "write anywhere file system" that allows for a flash copy algorithm that does not slow down as more copies are created. Instead of copying data into repositories when a write is committed, they write new data to a new chunk of disk and keep the old data where it is as long as it's still referenced by a flash copy. When you read from one of their flash copies, you get pointed to the original data, not the new one.
The problem is that while they don't slow down for many copies, they do slow down at higher usage percentages. It's a question of preference in the end.
I'm going to go read the full disclosure report to see if they short stroked the EMC box- it's standard practice to squeeze performance out of boxes in ways you would rarely see in real life in an SPC "bake off".

Archive User
Thu January 31, 2008 10:52 AM

Barry,
To answer your question, as long as the test will scale linearly you can scale performance linearly by adding RamSan-400 units to the array. It is increasingly common for our customers to deploy Terabyte arrays of our DDR solid state disks (Tera-RamSan).
I would be happy to have a more detailed discussion with you about Flash and the SPC-1 test offline.
For what it is worth, I continue to believe the IBM SVC shows some great performance characteristics and would not be surprised if the next generation doesn't beef up performance. If you get the SVC performance up, I think our Tera-RamSan array would be a great way to demonstrate that system's performance. I would also like to test the SVC in front of our RamSan-500 (2TB cached flash array). Woody

Archive User
Wed January 30, 2008 05:35 PM

Congrats to Woody and all at TMS for the new world record.
Be interested to see what you come out with in the Flash area as you hinted at over on my blog.
A serious question though, I presume the performance scales as you add more boxes (TeraSAN) as you'd need quite a few of these to get even 1TB of storage. Flash must work out a lot cheaper even at current enterprise prices to match capacity and pretty close to IOPs. Say 6x 36GB SSDs from STEC would be a lot cheaper, about the same IOPs and more capacity.
Still an impressive result and glad to see some more high numbers out there.
PS. I'm sure SVC will regain the crown with the next hardware base.

Archive User
Wed January 30, 2008 03:26 PM

Tony,
We accept the invitation into the Super High-End Club. Thank you for the invitation.
We would like to extend to IBM and others an opportunity to join us in the Super Low Price for Performance Club. Membership starts with sub $1 per SPC-1 IOPS results. Sadly, our club meetings are fairly boring sessions as we only get to meet with ourselves. If I did not think it was impossible, we would also invite other vendors to participate in our low latency club. Membership for this category requires sub one millisecond response times for all SPC-1 IOPS results.
Seriously, though, thanks for the blog and continuing to highlight SPC. I think it is good for the industry!
Woody HutsellEVP, Texas Memory Systems