Around the Storage Block Blog
Find out about all things data storage from Around the Storage Block at HP Communities.

EMC's VNX2 circus

CJZ Headshot fixed 150 x 150.jpgBy Calvin Zito, @HPStorageGuy  vExpert 2011, 2012, 2013.png

 

I've been blogging now for just over 5 years - and I've learned a lot about what readers want and don't want from a blog.   In my early days of blogging,  I did more than a few posts pointing out competitors' weaknesses (almost always a response to some misrepresentation they made about HP - Hi Alex!); those posts would get crazy traffic and comments.  Talking about a competitor is one of those areas that I usually avoid, preferring to talk about the good things I see happening in HP Storage. However, given the hype of EMC's announcement yesterday and some of the misleading  information about HP, I think it's a topic I have to visit.  But instead of offering my views, I thought highlighting what I've seen from a few independent bloggers and in the press would tell the story: 

 

Chris Evan's has a blog post with his view both on EMC's hype and the announcement itself.  He made several noteworthy comments:

clip-art-circus-465619.jpg

  • An EMC exec said, “We inserted flash into arrays designed for HDD and that’s what’s prevalent in the market today”.  Chris points out that this is not true - and mentioned HP 3PAR.
  • Another EMC exec said, “You can’t address the wide range of performance and capacity requirements with a single architecture” - again, Chris points out this is an EMC spin that I'd say is an outright lie.  EMC is well aware of the 3PAR architecture and the fact that it scales from entry mid-range, to tier 1, and our all-flash optimized 7450.  An interesting tactic by EMC to ignore the truth but I'm not surprised as I've seen this many times in the recent past.
  • Chris' conclusion is worth noting here:  "It’s difficult to get past the hype and see the real value here.  The VNX upgrade was well overdue but doesn’t leapfrog the competition and now we have some seriously mixed up sales messages.  VMAX and VNX overlap in their capacity and performance capabilities.  The new VNX-F would seem to be a competitor to XtremIO (which was conspicuously missing from the announcements).  Project Nile will add yet another storage platform to EMC’s portfolio, which now includes VMAX, VNX, XtremIO, Atmos, Isilon, VPLEX, Project Nile and ScaleIO, none of which have any similarities or interoperability.  Now you can see why ViPR is so important.  I said recently that EMC are becoming nothing more than a portfolio holder of disparate storage solutions.  The latest announcements reinforce that view.  As a business strategy this approach is clearly working because EMC continue to increase revenue with each quarter.  But does it actually serve the customer and help to deliver efficient solutions?  I don’t think so."

Another well known storage blogger who has weighed in is Martin Glassborow. He has a couple of blog posts - his initial thoughts on the announcement and then a follow up with one titled Don't Shoot the Messenger's Friends. A few comments Martin made that stand out are:

  • "It started off badly when they put the Sneak Preview video up and decided to have a dig at their lead competitor; then some tweets where they used the tag #NotAppy.

    And then the hype cycle started to ramp up. So we have a ridiculously overblown launch with some tenuous link to F1. Tyre changing competitions and the likes which appear to be fun but just break up the presentations and destroy the flow.

    EMC are just very bad at this sort of launch; speeds, feeds, marketing mumbo-jumbo linked in with videos/events which trash competitors, bore the audience and add little value. But all with the highest production values."
  • "Of course EMC continue to talk very fast about VNX being a unified platform when all reality; we know it’s not really…not in the way that NetApp is. But that’s fine but it still grates that Marketing smoke and mirrors are involved."
  • "I think if the announcements had been accompanied by blogs and commentary more akin to Storagezilla’s ....; much could have been forgiven and the announcement could have put to one side as the carnival it was."

A comment I saw from Greg Ferro sums it up well: "I also could do without EMC marketing style – it’s certainly egocentric and seems to be an attempt to build a marketing resume/portfolio instead of the company profile. Maybe that’s just me too."  I don't think it is Greg - this seems to be their pattern for announcements - infotainment type events light on info and trying heard to entertain.

 

Finally, in a Computerworld story I read that included comments from IDC analyst Ashish Nadkarni said this:  "The VNX still can't scale out as much as some newer platforms, such as EMC's upcoming Xtremio all-flash array, because it can only have two controllers. The Xtremio gets another controller with every module that's added to it.  "The VNX is, in a way, a legacy platform," Nadkarni said. "Within the context of a dual-controller system, I think it is a significant step forward."  EMC plans to keep both platforms going forward, with the less expensive VNX being an alternative to the faster Xtremio system, EMC's Herzog said."  EMC needs at least 5 platforms with what we can do with HP 3PAR - Polymorphic Simplicity!

 

A couple of other things I've heard that I think need to be said are:

 

  • Looks like existing VNX customers would have to do a forklift upgrade to leverage the new VNX2 software.  There's not a migration path that EMC discussed.  Ouch!
  • The core strength of HP 3PAR is it's ability to deal with performance in mixed and multi-tenant workloads. And most virtualized environments bring with it a mixed workload demand.  Despite all the high performance claims by EMC, I'm confident it won't do what HP 3PAR can in these environments. Calling it an array built for virtualization is a stretch. You can bet a lot of money that we won't see any SPC benchmarks from EMC.
  • Word is that the VNX2 software has millions of lines of new code.  EMC isn't known as the most innovative technology but always had customer tested products.  This isn't. And while I'm sure they have made some advances, it's still a legacy dual controller array.  Many of the updates that have been made are things that HP 3PAR has done since Day 1.

Bottom line for me is that every EMC VNX customer that is considering VNX2 should be at least kicking the tires with us and giving HP 3PAR a serious look.   I did a number of recent ChalkTalks that highlight the advantages of the HP 3PAR architecture that I think is great place to start.  Check them out!

 

 

 

 

Here's one more that focuses on the cache innovations specific to the all-flash HP 3PAR StoreServ 7450.

Labels: 3PAR| EMC| StoreServ
Comments
JohnMartinIT(anon) | ‎09-06-2013 03:52 AM

-Disclosure NetApp Employee-

 

Nice post Calvin,  classy effort, it's hard to write about a competitors launch without being too partisan and I've been wondering myself if I should weigh in on my own blog  (I've decided not to).  As far as ignoring 3PAR's contributions in the all flash performance space, I suspect that EMC's comments were mostly targetted at FAS/ONTAP but even if that was the case they got a lot of that wrong too.

 

I suspect the complexity of thier portfolio, and the nature of their corporate culture makes it hard for them to imagine , or perhaps admit that anything outisde of that portfolio might actually be superior to one or more of their current offerings. To be fair to them, they do a lot of innovative engineering, and they've got some great tech. The work their engineers have done with optimising CPU balancing for traditional flare LUNs is impressive, in the same way the work they did with their X-Blades in the VNX1 line was also impressive. I'm almost as impressed by that as I am with the work HP has done with 3PAR since they purchased that product line, and the engineer in my really hopes that theres more good stuff to come there even while the competetor in me wishes otherwise :-)

 

In the end though, I alsmost wonder why they bothered ... it's interesting from an engineering point of view, but I never really saw them say where the significant customer value was in having a gazillion IOPS in a dual controller box for old-school fixed RAID LUNS . From my point of view it, there's not a lot of point it if  doesnt also address challenges like capacity efficiency, fault reslience, planned downtime, application self service, and multi-tenancy and cloud integration  (just to name a few).


Then again, it could be just me, because even as much as I like to watch an F1 race, even in my teens I never really saw the point in trying to prove my manhood by racing other at the traffic lights, even when I had the much better car, it just seemed kind of pointless. 

 

Regards

John

 

| ‎09-06-2013 06:06 AM

Hey John - I wrestled with this one, a lot.  My original idea was going to be to focus on HP 3PAR with maybe a tiny bit of comparison to VNX2. But I obviously changed my mind along the way. Between this announcement and "breaking records", I think EMC marketing has jumped the shark. 

 

One of the ideas I originally was going to talk about here but I edited it out is that I don't think EMC has gotten where they are via technology leadership.  I've heard more than a few industry analysts flat out say they aren't innovative - they've had great marketing.  And as a long time marketing guy with an engineering background, I agree. I used to be in awe of their marketing machine.  But given what they've been doing lately - at least in their mega-announcement events - I think they're giving marketing a bad name.

 

Always good to hear from you - hope you are well.

Tom Callahan(anon) | ‎10-02-2013 01:19 PM

Good article, although obviously going to be based highly on opinions of the bloggers involved, and less around customer centric use cases. Every customer uses storage differently, tiers slightly differently, and has different and sometimes unique requirements.

 

I would like to point out, that you mention the VNX2 code cannot be pushed into the VNX, which appears to be true. However, in the 3PAR world, the same tends to be true since the feature sets are often defined based on which generation ASIC is in the array. And not just 3PAR, most storage vendors this will hold true, since obviously new generation arrays and features often require newer, faster, etc. hardware to go with them.

 

Thanks,

Tom Callahan

Leave a Comment

We encourage you to share your comments on this post. Comments are moderated and will be reviewed
and posted as promptly as possible during regular business hours

To ensure your comment is published, be sure to follow the community guidelines.

Be sure to enter a unique name. You can't reuse a name that's already in use.
Be sure to enter a unique email address. You can't reuse an email address that's already in use.
Type the characters you see in the picture above.Type the words you hear.
Search
Showing results for 
Search instead for 
Do you mean 
About the Author
25+ years experience around HP Storage. The go-to guy for news and views on all things storage..
Featured


Follow Us
The opinions expressed above are the personal opinions of the authors, not of HP. By using this site, you accept the Terms of Use and Rules of Participation.