back to article An EMC-HP Borg cube will totally ANNIHILATE its storage worlds

Suppose EMC and HP merged: the result would be an outrageous nightmare of overlapping storage products. Managers would run screaming into the darkness, and bean counters would spend millions converging products. It’s a no-go area, really. Quick and dirty look at HP and EMC storage products A quick and dirty look at HP and …

Page:

  1. Anonymous Coward
    Anonymous Coward

    Misunderstanding

    How is StoreEver, a line of tape libraries, remotely comparable to Isilon, a scale-out filer? To the limited degree HP plays in that space, it's with StoreAll, which we HP resellers refer to as Ibroke.

    1. Chris Mellor 1

      Re: Misunderstanding

      Arghh. Typo alert. StoreAll dickhead writer, StoreAll. Getting it fixed.

      Thank you!!

      Chris.

      1. Nate Amsden Silver badge

        Re: Misunderstanding

        While StoreAll is a scale out filer it does not compete with Isilon. HP re positioned it as an archive platform. Its not geared towards the same market as Isilon (HP used to try to sell it that way but it didn't work).

        StoreEasy may be their scale out filer for transactional workloads. Up to 8 controllers I believe and windows 2012 storage server software. Having used it for the past few months I am not too fond of it.

        StoreAll is also what powers they underlying storage of StoreOnce.

        HP offers helion openstack (downloadable) for the virtualization standpoint.

        I think hp data protector should go on that list perhaps next to avamar.

        HP arcsight and tipping point are both security products

        http://www8.hp.com/us/en/software-solutions/enterprise-security.html

      2. Anonymous Coward
        Anonymous Coward

        Re: Misunderstanding

        "it would have to junk the lot, apart from odd technology bits, as converging things like VNX and 3PAR would be a total and utter clusterfuck."

        Really? who in their right mind, even in EMC' universe would want to continue down the VMAX, VNX, VNXe, VPLEX, XtremeIO...... route when they can have a single, modern, feature rich platform do the lot !

  2. Trevor_Pott Gold badge

    Since when has EMC given fucks about the politics of the situation? The EMC federation is a terrifying morass of backstabbing and fifedom-guarding. VMware has been repeatedly described as being a political "thermonuclear wasteland" and EMC isn't heaps better.

    I believe you underestimate the egos of the people involved. They honestly think they can just "win" because they're that much "better" than everyone else. All evidence to the contrary. The EMC federation is not where sanity or rationality rest for long.

    1. Anonymous Coward
      Anonymous Coward

      Going to Pott

      Ex-EMC employee here.

      Was your mother frightened by an EMC salesman while you were in the womb?

      I can't deny that EMC is very political, sometimes intensely so. But it was a mostly very pleasant place to work, markedly much more civil than most of the startups I've ever worked for.

      Take a chill pill, dude.

      1. Trevor_Pott Gold badge

        Re: Going to Pott

        "Was your mother frightened by an EMC salesman while you were in the womb?"

        Nope. I couldn't give a bee's bollocks about EMC. Not my job to care. It is, however, my job to investigate. To ask questions others don't want asked and to root and the answers people don't want told.

        Interesting nerve I touched...

    2. Mpeler
      Flame

      I hear the sound ot two garbage trucks colliding....

      To quote Scott McNealy's comment about HP merging with Compaq.....

  3. P. Lee Silver badge

    Mergers Happen Because

    Execs want to pad CVs

    Execs want get bonuses without improving the business

    Execs want to hide losses or blame someone or something else

    Merging reduces competition

  4. Anonymous Coward
    Anonymous Coward

    Just a thought...

    Perhaps HP fears that EMC would fall to the hands of Cisco?

    Then again, where does this HP-EMC rumour originate, fortune cookies?

  5. Michael Duke

    SMB Array - EMC has VNXe

    HP has a similar capability to VPLEX in Peer Persistence on the 3Par platform although less flexible and more feature limited, however a good match for VPLEX Metro when used with VMware or Hyper-V

    1. Anonymous Coward
      Anonymous Coward

      HP's 3PAR platform actually covers more than you suggest and there's nothing to stop it going further up or even down the stack if required.

      3PAR 10000 covers VMAX in all areas but mainframe. 3PAR 7200 & 7400 cover midrange and high-end VNX and the 3PAR 7450 covers Xtremio. A single platform based on a modern architecture with common management, both baked in and mature feature set etc. Like the other poster said all the 3PAR's also support peer persistence, which also covers a huge part of Vplex's market.

      I wonder who would win that battle?

    2. StorageWorldCitizen

      EMC VPLEX matches to HP XP7 Active-Active

      In my condition of HP storage salesman, I just want to add some details to Michael Duke's comment:

      3PAR Peer Persistence (it is a SW feature, not a virtualization appliance) can match to VPLEX when talking about VMware vMSC Active-Active scenarios, but they are absolutely different animals. The right matching is VPLEX vs XP7 Active-Active, wihich is the HP OEM naming for HDS VSP g1000 storage virtualization gateway.

      1. Anonymous Coward
        Anonymous Coward

        Re: EMC VPLEX matches to HP XP7 Active-Active

        I agree from a hardware perspective, but disagree on the functionality, ~90% of what VPLEX does can be covered by Peer Motion, Peer Persistence and CLX in a HP 3PAR environment. The remaining 10% i.e corner cases are covered by XP7 Active / Active.

    3. StorageWorldCitizen

      EMC VPLEX matches to XP7 Active-Active

      Just to add some detail to Michael Duke's comment:

      3PAR Peer Persistence (it is a SW feature, not a virtualization appliance) might match to VPLEX only in VMware vMSC Active-Active architectures, but they are absolutely different animals. The right matching is VPLEX vs XP7 Active-Active, wihich is the HP OEM naming for HDS's VSP g1000 storage virtualization gateway (in case you ask, I honestly think it is much better than VPLEX).

  6. thondwe

    Major benefit of Cloud

    Major benefit of Cloud is that someone else has to deal with the fall out of these sorts of mergers! Is it me or is has the Storage market got far too many players in it now?

  7. Anonymous Coward
    Anonymous Coward

    Btw EVA Is an end of life product replaced a long time ago by 3PAR, it's no longer for sale so shouldn't be on this list, unless you want to include EMC CX also

    1. Anonymous Coward
      Anonymous Coward

      EVA is still a current product, sold alongside 3PAR.

      You can still buy new EVA or add to your existing EVA arrays.

      http://www8.hp.com/us/en/products/disk-storage/product-detail.html?oid=5062117#!tab=models

      1. Anonymous Coward
        Anonymous Coward

        You can still buy EVA parts and spares but the products have been formally end-of-lifed for a long time now. If they don't show up that way on HP's website, that's HP's demonstrable incompetence not strategy.

      2. Anonymous Coward
        Anonymous Coward

        Just because it appears on the website doesn't mean it's a current product, I'm sure the same is true for many other vendors discontinued products.

        EVA or more precisely P6000 went end of life in January 2014

        http://www.atr.si/f/docs/Aktualno/EVA_P6x50_EOL_Customer_Letter.pdf

        Of course you can still buy parts and obtain support and that will continue for a minimum of 5 years beyond the end of life announcement.

  8. Anonymous Coward
    Anonymous Coward

    Actually 3PAR and StoreServ are the same platform (3PAR StoreServ), maybe a bit more due diligence was In order around HP's product set and capabilities.

    Remove the legacy end of life stuff, the above double counting, get the positioning right and it's a much simpler and more capable portfolio than is being suggested.

  9. Smoking Gun

    I'm surprised there aren't many rumours of Cisco sniffing about. When you look at Cisco's datacenter play you would think EMC was an ideal fit but Invicta and annoying NetApp.

  10. Anonymous Coward
    Anonymous Coward

    Technically, EMC has a public cloud, if you count VMWare

    vCloud Air (formerly known as VCHS/vCheese).

    I believe you reviewed it favorably on ElReg just a few weeks ago.

  11. Anonymous Coward
    Anonymous Coward

    Place your bets.. mine are:

    Winners after battles:

    ------------------------------

    - VMAX (what are HP's margins & IP on an OEM'd product ?).

    - 3Par (VNX will put up a hard fight for sure).

    - VNXe (over MSA, which is from an aging DEC portfolio)

    - Documentum (Nobody likes Autonomy anymore)

    Winners by default

    ----------------------------

    - Vmware (no real option from HP)

    - Isilon ("")

    - Pivotal ("")

    - DSSD ("")

    - RSA ("")

    - VIPR ("")

    - VPLEX ("")

    - VSE/VSPEX ("") (*but withreduced cisco love?)

    - All of HP tape products (nothing from EMC here)

    - Helion ("")

    ------------------

    Co-exist by segmenting markets:

    - XtremeIO & 3Par AFA

    - DataDomain / StoreOnce / Avamar

    - Atmos / StoreAll

    - ScaelIO/ Storevirtual

    -----------------------------

    Losers:

    - Centera

    - VSP

    - MSA

    - Autonomy

    ----------------------------

    Only point of serious contention seems to be VNX and 3Par, which 3Par may win.

    1. Anonymous Coward
      Anonymous Coward

      Re: Place your bets.. mine are:

      The problem with 3PAR is a lack of unified capability - they've been trying to hack Linux NAS into it for some time and when it's introduced it'll be more of a mess than VNX NAS (even if 3PAR block is superior).

      StoreAll is going to bite the big one, merger or no merger, and soon. HP isn't getting in bed with makers of working object systems for nothing. IBRIX is a failure and an embarrassment.

      MSA sticks around because HP can buy it really cheap from Dot Hill and it's a reliable, performant entry-level SAN array. Cheaper than the equivalent long delayed and probably stillborn low-end 3PAR for sure and possibly more so than VNXe.

      VNXe is really cool, but it'll be hard to sell just VNXe without VNX, as hacky as the latter may be. However don't forget it's highly functional and reliable despite a lack of inner beauty.

      1. Ant Evans

        Re: Place your bets.. mine are:

        Spectacular and imminent destruction of wealth through ill-advised mergers. You have been warned.

        The dynamic is similar to IPOs. Sweet-talking matchmakers bustle selflessly around the shy and vulnerable (IPO) or fat leftover (M&A) people, create a match made in heaven, and then scarper with their 10%. Relationship counselling not included. No refunds. Now fuck off.

      2. Anonymous Coward
        Anonymous Coward

        Is that cool VNXe still window storage server under the covers ?

  12. Dapprman

    Weren't similar things said before the merger of HP and Compaq ?

    Just saying as I have no loyalty to either HP nor EMC. Sure it could be a disaster, but HP have been there before and come out quite quickly smelling of roses.

  13. hoola

    Ageing Platform

    EMC have an ageing suite of storage that is layer upon layer of software on ancient hardware in an attempt to keep up with the competition. It only sells because management see it as a "safe bet". If EMC do not do something to bring the stack into the 21st century they are toast. There are so many better solutions out there they only way they can sell is by paying (bribing?) Gartner to get into the correct quadrant and to do a hard sell to non-technical staff.

    The HP 3PAR solution walks over any comparable EMC product. The low end MSA is trusted and rock solid. Isilon and ATMOS will be eaten alive once a full Object Store stack comes from one of the forward-looking, current niche players (DDN as an example).

    1. Anonymous Coward
      Anonymous Coward

      Re: Ageing Platform

      Probably also worth mentioning 3PAR has built in online migration from VNX storage............

    2. Anonymous Coward
      Anonymous Coward

      Re: Ageing Platform

      There's nothing ancient about EMC hardware...software maybe. If you want ancient hardware, look at the long neglected "high-end" 3PAR or things running on old Proliants.

      1. Anonymous Coward
        Anonymous Coward

        Re: Ageing Platform

        "There's nothing ancient about EMC hardware...software maybe. If you want ancient hardware, look at the long neglected "high-end" 3PAR."

        If you've been listening to EMC for the past 5+ years they're apparently all about software, yet by your own admission EMC's software is ancient.

        The 10000 series hardware was designed and implemented before 3PAR had access to HP's supply chain as was the F-Class, before it was replaced by the StoreServ 7000 and look at how that's turning out !

        Yet still the ancient old decrepit 10000 series has the ability to compete and win at the top end of the market and that's because in the real world the hardware doesn't really matter (it's all off the shelf components anyway). The software and secret sauce make the real difference and that's what 3PAR has in spades and EMC doesn't.

        Hardware refreshes are relatively easy it's just a matter of timing, but if you aren't adding the secret sauce then you aren't going to be able to take full advantage meaning shorter life cycles and higher TCO for your Customers, but maybe the marketing dept can help out with that :-)

        1. Anonymous Coward
          Anonymous Coward

          Re: Ageing Platform

          Based on the commentary between, anon-cows from HP and EMC, looks like this merger is going to be extra fun...

        2. Anonymous Coward
          Anonymous Coward

          Actually EMC's hardware is relatively up to date (coutesy of Intel) as is a large proportion of the outer software layers running on their platforms. When you compete with EMC they like to concentrate on the latest and greatest hardware as a differentiator. Typically they pick a single on paper metric in which they have a numbers advantage e.g CPU clock speed, cache size etc and hammer it for all its worth to the customer, despite the fact that architecturally they're comparing apples to oranges.

          The big problem is that the core software that forms the underlying architecture of these platforms is still in place. EMC have been lazy over the years and decided to just layer on functionality in response to market pressure instead of fixing the underlying software architecture. The fact the underlying software architecture is now decades old now limits everything they try to do, that's why they had to go out and buy XtremeIO from a startup as they had no way to make VMAX or VNX competitive in the all flash market.

          But rewriting code from the ground up means many millions of upfront investment, years of field time are required to get the maturity and subsequent availability. If further proof were needed witness EMC's latest gaff with XtremeIO data destructive upgrades. You also need to treat customers with kid gloves providing long term roadmaps and a simplified migration method to keep hold of them during the transition. So why slow down the gravy train when customers are gullible enough that you can paper over the cracks with slick marketing.......... the problem is that gravy train is now starting to run out of track.

  14. Anonymous Coward
    Anonymous Coward

    Most of those HP products would probably be migrated, or EOL, into the EMC lines because they have small market shares today. The reason HP would want to merge with EMC is to get VMware, not storage.

  15. Anonymous Coward
    Anonymous Coward

    bloody storage streets

    "It’s fantasy that would result in blood on the storage streets."

    Oh, well then, let's hope it happens.

  16. abufrejoval

    Imagine the Memristor was real...

    Imagine for a moment, that what Meg Whitman and Martin Fink announce in the opening session of day 2 of HP Discover might actually come true:

    Imagine the memristor is real: Hard disks, flash and all DRAM up to L3 cache will no longer be able to compete with memristors on performance, energy consumption, physical density and price per unit of storage.

    Meg Whitman literally says she is betting the company on this happening.

    Now imagine an EMC without spinning disks and flash.

    Imagine also an HP without any of those.

    What both will still need is customers and those customers will need to run their software and services to make money and pay HPEMC to make it happen.

    HP is trying very hard to make the transition for those customers as easy and painless as possible and that may even include *not* lowering their prices as fast and as sudden as the memristor technology might be able to (truly heartwarming that!)

    Parts of EMC are working very hard on creating software assets, which manage compute workloads, geographically dispersed policy managed data and software defined everything.

    Those parts and the customer base of HPEMC is what is attractive to the enterprise part of HP.

    The rest will follow the vacuum tube and core memory.

    If you can spare the time, please have a lookt at the full HP discover session:

    (youtube prefix) watch?v=33WPphbFeDY

    And then pay close attention to the joint session (starting ~54:04) with HP, Intel and Microsoft, which is two old time HP bedfellows who are very carefully prodded by Meg Whitman to either follow along, or join the vacuum tube. You can see how both of these senior execs feel somewhat uneasy at being offered a junior partner role, while their brain should be telling them that it's the best they could hope for, when HP has an exclusive technology lead which makes the AMD/Intel gap look miniscule.

    Intel clearly isn't as much into SoCs as HP would expect them to be and Microsoft ist mostly about customer base and protecting HP revenue flow.

  17. klaxhu

    Reading these comments is good fun

    Oh ..so the consensus here seems to be hp's products will come out winning in a potential merger ?(very improbable btw) having worked both at hp and emc, I can only say this is a load of bs and clearly people are expressing opinions without having a proper insight.

    No matter what each of the companies might tell you as an employee or partner, the numbers talk and the rest is fluffy talk: EMC federation on the up, HP overall on minus.(36% EMC, under 10% for HP marketshare and these are only storage numbers!!!) Why the merger would make sense has totally other reasons than what products to keep and which to kill.

    To take one example: when you compare an aging 3PAR all flash platform vs a build from the ground up AFA like XtremIO which competes with real AFA's in the market like Pure ...clearly you don't understand the product. Plus, EMC federation's big bet from now on is software and cloud and not hardware as you can read on all EMC-ers blogs or from the messaging in the media as well ..so: VIPR, vCloud Air, Pivotal (for PaaS just as an example).

    My 2 cent: XP (its Hitachi OEM, so another reason to kill the partnership: hp been selling 3PAR to ex EVA/XP customers anyway to replace them so a no brainer there). 3PAR definitely win over VNX just because of the multiple controller architecture. VMAX will stay for mainframe and the high end resilience platform for now. Lower end VNX/e. All software products can stay and offer choice. Backup and file on Datadomain and Isilon.

    The big ones are: will VMware be spinned off or Helion used or a mashup between the two? What will happen to Pivotal and RSA? (you can't really compare some of the security products HP has to the big RSA machine! be serious ...they have hundreds of products that sell really well, how much revenue do HP do from their security division ?)

    Let's not forget that EMC has the same market cap and revenue power with a fraction of employees vs what HP Ent does as a whole today.

    EMC's big positive is the sales organisation & support one which is unmatched (hence so many banks use them). Guess that is why you see no 3PAR in banks nowdays

    1. Anonymous Coward
      Anonymous Coward

      Re: Reading these comments is good fun

      "To take one example: when you compare an aging 3PAR all flash platform vs a build from the ground up AFA like XtremIO which competes with real AFA's in the market like Pure ...clearly you don't understand the product."

      That would be the XtremeIO box that has one of the worst footprints in the industry, lacks even basic features such as online expansion, online upgrades, requires data to be migrated off to take the new firmware, has an external UPS system for cache protection and needs a speed bump between it and the host in the form of Vplex to do anything even remotely interesting like DR.

      Both the 3PAR 7450 and the Pure wipe the floor with Xtreme in any serious test, so I doubt those many banks you crow about will be wanting to implement a beta product anytime soon

      1. klaxhu

        Re: Reading these comments is good fun

        Do you have any proof to support that? I don't know any bank (in the UK or the US that is) that uses 3PAR AFA or Pure. XtremIO however, is another story. "wipe the floor" is a very generic term - can you show some real life examples? Cause controlled environment benchmarking ..we know how good HP is at that.

        There is a reason why EMC never takes part in these kind of "public" tests so curious where you get your info.

        Not saying 3PAR is a bad product. It's great, ..just nothing like Pure or XtremIO. Again, the numbers speak for themselves: you put some empty sentences together with no actual facts or official shipping numbers.

        If we look at the numbers, EMC is still shifting the most Flash in the storage industry.You can't fault numbers! Unless you forge them of course

    2. Anonymous Coward
      Anonymous Coward

      Re: Reading these comments is good fun

      For once Gartner don't agree with you / EMC, they place the 3PAR P10000 as pretty much neck and neck with VMAX and the HP XP P9000 above VMAX in every category.

      http://www.gartner.com/technology/reprints.do?id=1-1RO1Z8Z&ct=140310&st=sb

      1. klaxhu

        Re: Reading these comments is good fun

        Wow, good! yet ..again, nobody buys the bloody stuff in banks. They prefer either the clean VSP from Hitachi or the VMAX in banks... for example. And I agree, VSP is a great product as well as VMAX is.

        Yet, we were not talking about this were we? You mentioned some floor wiping of the 3PAR ...

        Since you love Gartner and IDC so much, why don't you share their latest creation in the AFA space?

        Here you go, I will do it for you:http://www.gartner.com/technology/reprints.do?id=1-20P2W9A&ct=140902&st=sb

        I know I know ..you must be thinking now: EMC must have paid them ...but then that would make your link invalid as well, would it not ? :-)

        It's funny how you have no clue about the facts, yet somebody must have filled your brains with FUD.

        See ..I have been there working for both, and I remember well how HP used to brain-wash us daily to say we are the greatest and first in the market when the numbers were saying something else.

  18. Anonymous Coward
    Anonymous Coward

    Wow,,,just wow....I see you didn't even attempt to address any of the XtremIO weaknesses, best to just ignore those and move on and hopefully they'll get lost in the noise eh ?

    I would hope EMC agree that Customers running a POC for an AFA product should really test more than just the all out speed of the box. Yet outside of that rather narrow measurement XtremIO just isn't competitive, it lacks not just enterprise but basic features and putting a VPlex in front to address those challenges is just going to increase your latency, it will also lower your IOps whilst increasing overall complexity and cost.

    Now you mentioned FUD, so please if any of the above is then please enlighten us as to which bits are FUD.

    Lets not pretend EMC has some kind of exclusivity in the banking and financial services sectors, 3PAR has plenty and already had a few even before HP made the acquisition and yes there are also 7450 AFA's going into those and net new accounts, but that's OK I'm sure HP haven't been keeping you up to date on each deal. Not sure about Pure but I see them everywhere at the moment mainly banging on the door of EMC accounts, but given the noise they're making and EMC's panicked response I'd assume they're having some amount of success.

    I had to laugh at your suggestion HP somehow engineer benchmarks in their favor. The irony of this coming from EMC who A. won't publish benchmarks, B. tie their own Customers in so many T&Cs that they can't publish results themselves and C. are brazenly touting a benchmark kit to Customers specifically engineered to make XtremIO look good, sometimes the truth is much stranger than fiction.

    Since you brought up the AFA report from Gartner it's probably worth mentioning that EMC are heavily reliant on their position in that MQ on their legacy VNX-F rather than some miracle performed by the release of XtremIO. Have a look at the opinions of that particular MQ on some independent blogs outside the EMC cheer leading crowd and and you'll see it's seen as pretty useless as the market just isn't mature enough yet.

    If you really want to compare then you should, be looking at the Critical Capabilities Report not the Magic Quadrant but then that would expose all those XtremIO shortcomings all over again wouldn't it ?

    "I remember well how HP used to brain-wash us daily" Touche I heard exactly the same from an ex EMC guy only a few weeks ago. So extreme was his belief that it took him several months to make it out of denial :-)

    1. klaxhu

      Sorry, ..are we looking at different charts/links?

      Open again my link and tell me if you see written there XtremIO or VNX-F in the AFA leader quadrant? Or maybe you need some glasses. Also ...do you make the difference between an array like VNX or 3PAR in which you shove a lot of flash drives (aka Hybrid) and a AFA like Xtrem or Pure? Look again at that damn Gartner graph and see who the innovators are ...is neither EMC or Pure dude!

      You mentioned a beta product - what product are you talking about: again, look on the EMC website. XtremIO is GA for more than a few months now and on microcode 3.0 now.

      First you say it doesn't compare in performance with the 3PAR or Pure, then that it's basically nonexistent in the AFA market ..and you keep going...without presenting facts

      I shared with everyone the facts and your FUD just has no basis.

      Take a chill pill ...

      The difference between me and other is that I use my brain when I get hit with FUD, brainwash material, etc and also some logic and some google to search for numbers and facts (on vendor websites).

      You, par example ...you SIR ...are just spreading meaningless words with absolutely no backing here.

      I get it.. 3PAR is the best! If I would not known better, you either work for HP or are some 3PAR biased blogger ..and I know a few of those.

      Benchmarks: only real life data counts, benchmarks in a controlled environment are useless. POC a 3PAR vs a VNX and a VNX wins hands down ...I have seen it too many times. EMC don't even need to bring out the big guns (XtremIO). XtremIO is a real AFA and competes with the real AFAs out there.

      The enterprise features you talk about are in the software nowdays or in in things like recoverpoint or vplex..nobody cares about arrays can do. The world is moving towards host replication and application resiliency rather then everything in the hardware layer (see what the cloud providers are doing for example).

      The world is changing constantly ..and clearly, you are not keeping up with the trends.

      1. This post has been deleted by its author

      2. Anonymous Coward
        Anonymous Coward

        I took a look at the Gartner link, if you read beyond the graph you can see the rating was based on a combination of VNX-F and XtremIO as discussed in the Vendor strength and cautions.

        "EMC has two SSD-based products in the SSA market: (1) the XtremIO scale-out technology, which EMC acquired in May 2012; and (2) the VNX-F array, which is based on the traditional general-purpose VNX unified storage array".

        1. Anonymous Coward
          Anonymous Coward

          Thanks for the confirming what I already wrote, I see you talked a lot about FUD but didn't actually deny or address any of those points. BTW I wasn't here to bash EMC, just to comment on the potential merger, but your half baked commentary makes it almost impossible not to.

          VNX wins hands down eh ! I've seen many real world Customer POCs now where 3PAR pretty much destroys VNX across the board, BTW you asked for that ! I know you're not going to admit it but take a look at your numbers and you'll see something beginning with 3 and ending in PAR is taking market share directly from VNX, it's also the only product growing in the high end where VMAX plays. I'm not saying VNX is a bad box but it's limited by the underlying architecture, so putting it up against a true active / active array with a fully virtualized back end like 3PAR, amounts to little more than a fools errand. If you can't see that then you're beyond redemption.

          Now 3PAR has inline dedupe baked into the product and it also has a industry leading cost per usable GB @ $2 for an AFA, adaptive sparing and caching for solid state, sub ms access @ 900K IOps, great power, cooling and rack footprint and a full complement of enterprise features. I'm not claiming the 3PAR 7450 has the largest market share, is the fastest on the market or has the best dedupe ratios etc but if you look holistically it's more than fast enough for >90% of use cases (and much faster than a lot of it's AFA competition) and also is the most feature complete with a field proven enterprise code base, so what's not to like vs the one trick ponies out there.

          It's worth mentioning that Gartner's ratings didn't take dedupe into account on 3PAR because of the timing of the report and I doubt very much whether they factored on XtremeIO's inability to perform non data destructive upgrades, so I think we can take the current ratings with a pinch of salt. As for the 3.0 microcode the versioning is laughable given how long the product has been GA and the fact you can't even perform an online upgrade. I see when challenged you also brought out the old chestnut that EMC use as a get out of jail card for not publishing benchmarks, "because they're not real world", highly ironic given the benchmarking kit EMC are pushing at XtremIO Customers as part of a POC.

          Ah and of course everyone wants to buy multiple products and cobble them together to address basic array functionality and those large environments love having to manage all those features on a per host basis. If that were the case then why is EMC telling Customers they'll have these features baked in in the future, why not just rely on all the other platforms or the host O/S to fill the gaps. Because they know the current position if unsustainable in the long term and others who have these features will eat their lunch.

          Thanks for calling by.

          1. klaxhu

            The fact that you even mention something from the marketing slides like @900 IOPS without mentioning a real life workload determines you have no clue what you are talking about and you haven't actually seen any POC's. It's nothing new that 3PAR or Netapp steal share from VNX and vice versa. For every 1 customer that 3PAR wins over VNX, EMC wins 4 other customers. Just look at the numbers ..again :-)

            However, you are right - let's focus on the merger potential of these two companies. Knowing both well, I am pretty sure it won't happen, but if it does, a lot of HP (yes, not EMC) employees will lose their jobs.

            What products will survive? Clearly the ones with the bigger footprint within their market, and maybe a few with potential, but that's it. EMC has the best product for every type of workload (which if you ask my personal opinion creates quite a lot of products and it's complex!) where 3PAR or Netapp ..or even Pure have just 1 which they try and make it work for every type of workload.

            Hardware is commodity as margins prove it, be it networking, server or storage. The money is in the software (just look at oracle's or ibm's strategy or lessons within the tin layer). If you job is in knowing the technical bits of what an array can or cannot do I suggest you smart up as you won't have a well paid job for long...

            1. Anonymous Coward
              Anonymous Coward

              Again you've failed to address any of the questions posed instead choosing to throw more FUD and insults. Specifically what were those facts, numbers and references you've shared earlier ? Because despite rereading your posts several times, I can't seem to find any of them !

              "you have no clue what you are talking about and you haven't actually seen any POC's."

              Really it just so happens I and many of my colleagues have run a large number of POCs for spinning disk, hybrid and flash arrays and many of the former against VNX (its part of my job) and the latter now against XtremIO so don't attempt to suggest I don't understand this process. It's specifically why I highlighted earlier that a POC is not just about all out performance and hence why XtremIO isn't competitive in any properly defined POC. Given your obvious ignorance of the POC process this suggests to me that actually you're the one lacking in experience in this field .

              Ah yes marketing numbers, something all of the flash vendors are publishing, please show me some real world, audited numbers for XtremIO with IOps, latency measurements and workloads clearly defined and then we can compare. Yes I know, not gonna happen, so until then, all the other vendors will stick to lab numbers just the same as EMC.

              "It's nothing new that 3PAR or Netapp steal share from VNX and vice versa. For every 1 customer that 3PAR wins over VNX, EMC wins 4 other customers. Just look at the numbers ..again :-)"

              A tacit admission there then on the market share front ? followed by a statement that really doesn't make any sense whatsoever, do you have an infinite pool of Customers ? Yes I've looked at the IDC figures again worldwide and regional and guess what VNX is losing market share, maybe you need to take a look assuming you have access to such material which I doubt.

              "Where 3PAR or Netapp ..or even Pure have just 1 which they try and make it work for every type of workload."

              Single product eh, that's why HP have MSA, StoreVirtual, 3PAR and XP with MSA providing commodity and DAS, StoreVirtual software defined and 3PAR spanning mid range through AFA to the high end and XP for the traditional big iron market. HP aren't saying 3PAR is the only choice, but it does fit and fit well for a huge part of the market. Even Netapp have multiple platforms these days, so your wrong again.

              EMC can't hope to address the same range with a single product line and the reason for having so many platforms isn't a technical decision designed to benefit Customers in some way, it's because they have consciously decided to not integrate or extend any of their acquisitions over the years. Hence why they need VIPR to paper over the huge management and functionality chasms between product lines.

              I'm not too worried about commoditization, I've worked across a broad portfolio of products and categories in my years, so no need to smarten up. But even if software defined eats the array market, basic I/O and storage principles will still remain, even more so as you're never entirely sure what's under the covers, it's just the tin got cheaper. So I figure I have a job for a while yet, although it seems EMC could be the one in a bit of a hole going forward given their over reliance on high margin big iron.

              Finally you may believe I'm biased, but your own opinions aren't looking particularly independent either, at least I've attempted to explain my points, whereas yours remain highly defensive, light on any factual content, despite repeated statements to the contrary, whilst bordering on the aggressive.

              Maybe you need to lay off the Kool-aid a little....

            2. Anonymous Coward
              Anonymous Coward

              Back to the merger :-)

              IMHO HP would probably only want VMware and a few odds and sods, the rest of the technology, given software defined is apparently the future :-) and the distraction of trying to integrate the product lines wouldn't be worth the asking price.

              Even if they managed to retain some of the Customer base through the transition period, every other storage vendor would be having a field day during the process.

              Having VMware in their back pocket would give HP access to pretty much all of that Customer base anyway although they'd have to have a similar hands off approach to EMC. But really I can't see it happening, which is a shame given the potential (ugh) synergies between VMware and the soon to be HP Enterprise.

              Looks like it's off anyway, HP appear to have resumed their share buyback. That still doesn't take the pressure of an EMC / VMware split if the activists get their way though.

Page:

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019