2.044: ibm dumbs down storage marketing (again)
OK, this isn't going to be another one of my competitor-bashing diatribes. I've learned my lesson, based on reader feedback on my comments about IBM's past transgressions (who can forget IBM's initial "Let them use Tape" response to flash drives?)
Nope, this time, I don't have to be the one to do the tear-down: independent storage consultant self-proclaimed IBM proponent "PRJ" has exposed the (dare I say it) stoopidity in his post IBM Storage UK Has Codified Stupidity. In the middle of his post he highlights the following.
Yet again, this does not mean XIV does not meet some needs. What it does mean is that XIV is still not equal to nor does it offer performance comparable to the DS8000, and that IBM has said you - the customer - are too stupid to understand this blatantly obvious fact.
If I wrote that, you'd have blasted it as blatant FUD. But this guy says that he LIKES IBM…go figure.
And it appears that IBM Storage US is no better.
Case in point: in his latest post covering this week's IBM Storage product announcements, arch-nemesis Tony Pearson couldn't resist taking an entirely unrelated swipe at me and V-Max at the end of his post. (Tony clearly didn't appreciate my publicizing the impending death of the DS68000, nor my chastising of the way he (apparently intentionally) twisted a recent Chuck Hollis post into the offensive and insensitive accusation that EMC markets storage to terrorists).
In his attempt to take the high ground, TonyP steps into the land of Codifying Stupid when he includes a link to an (IBM-funded) "ITG white paper" titled Cost/Benefit Case for IBM XIV Storage System - Comparing Costs for IBM XIV and EMC V-Max Systems, and he then uses that paper to support an assertion that the XIV is up to 63% less expensive than "a comparable" V-Max.
With an assertion like that, you know I had to respond.
a white paper shouldn't insult the intelligence of its readers – imho at least
This white paper couldn't be more biased if IBM had admitted that they wrote it themselves. It doesn't take much digging to realize that this white paper is anything but objective, and that it clearly compares apples to steamships:
- The paper was clearly funded by IBM. In fact, ITG itself seems to exist largely based upon such funding from IBM, judging by the number of IBM/ITG white papers and references you'll find on Google. The up-front disclaimer even says (in effect) that the paper may not reflect reality. No Duh!
- The paper "admits" that there was an insufficient number of V-Max users at the time of the report (3 months after the V-Max announcement), and so instead admittedly derived
muchalmost all of the V-Max comparisons from customers' DMX experiences.
- Administrative overheads where derived from DMX - Given that V-Max has delivered comprehensive improvements in the simplicity and efficiency of storage management over the DMX family, this is an abject misrepresentation of the V-Max. ITG also did not consider the added storage management complexity required to support customer-required capacities above 79TB; XIV needs multiple arrays to do what V-Max can do with a single array (and quite a few to deliver the 2PB+ usable capacity of today's V-Max).
- Performance comparisons were also derived and not tested – in fact, most of the performance comparisons presented in the paper occurred BEFORE the V-Max was even announced! At one point, ITG asserts that they have no reason to think the V-Max delivers more performance than an XIV array, and thus made no adjustments for performance. I don't think I have to explain how silly that statement is.
- Although positioned as equivalent comparisons, the V-Max systems were not actually configured similarly to the XIV arrays in the comparisons. The XIV arrays offer only thinly-provisioned, wide-striped 1TB SATA with (effectively) mirrored protection; the V-Maxes were configured without Virtual Provision and using only "FAT" provisioning with 300GB 15K rpm drives. A more equivalent comparison would have been to also use mirrored 1TB SATA on the V-Max with Virtual Provisioning to get the benefits of thin and wide striping – plus a significant reduction in storage allocation complexity. The V-Max could have been configured with RAID5 for even more savings, but the fictitious comparisons that ITG invented avoided using any of these cost-saving efficiencies of the V-Max – without explanation.
But the biggest farce of the paper is the very notion of comparing the XIV to a V-Max in the first place. The two arrays are as different as are the XIV and DS8700, for pretty much the same reasons that PRJ points out in his blog. The spec sheets alone show even the most casual reader that these two systems don't fulfill the same requirements.
Like PRJ, I'm not calling TonyP an idiot. He probably didn't take the time to objectively read the paper he referenced, and he probably didn't really mean to insult the intelligence of his readers.
But just as PRJ said of Steve Legg, TonyP "basically told the world that he thinks IBM’s customers aren't smart enough to spend a few moments reviewing a spec sheet [or a white paper], and seeing the obvious disparity between the two arrays."
I couldn't have said it better myself.