NetApp: Some good and some errrrr…..

By | August 6, 2010

So there’s a bit of a storm rumbling through as a result of some statements that the Three Amigos (@storagebod, @ianhf and @chrismevans) have recently made.

So here’s my 1000000% unbiased independent <cough cough> opinion on some of the things being said.

One comment doing the rounds is “Large enterprises are stretching NetApp’s technology and they’re not keeping up”

I can;t see any honest person having a problem with that  statement other than may be the fact that truly large enterprises are stretching all of their vendors technologies.  Let’s face it, all of the vendors are struggling to cope with the explosive growth we have seen over the last howevermany years. 

I also don’t think there is an honest person out there, in the know, who would argue against the statement that there is a ton of NetApp Sprawl out there in the large enterprises – and credit to NetApp for selling so much kit and cornering the market.  However, this sprawl does cause problems, and such estates are a nightmare to manage.  In fact, more often than not they actually don’t get managed.  Then, when it gets really bad, those large enterprises end up calling in people like Chris, and some of the rest of us out there, to tidy up the mess. 

DISCLAIMER and honest fact:  I know Chris, and I know he is speaking from experience on this subject.

THOUGHT: One commenter even went so far as to call Chris "dangerous".  I therefore propose that Chris change his name to Chris Danger Evans – a'la Austin Danger Powers!!!

If people don’t think that large enterprises are pushing NetApp technology beyond what it was designed for then those people almost certainly don’t work for large enterprises, or they are lying.  Just look at the Spinnaker purchase (way back in 2003) and what NetApp are trying to do with OnTap 8 Cluster Mode.  They’re not going through the pain of making those changes just for the heck of it.  They’re doing it because they need to.  In fact somebody may need to light a fire under their arses because they’ve been sitting on that Spinnaker code for a long time now and scale-out NAS is well and truly here…. . 

I also defy any honest person to disagree that OnTap 8 with its two modes (7-mode and cluster-mode) is anything other than a rats nest.  I mean how complicated do you want to make things for your customers!?  With so many half-baked features it does little more than cause confusion.  Could they have done this any worse?  However, they could come good on it over the next few releases.

 

 

Balancing it up

However, even if NetApp technology is being pushed and arguably not keeping up with the huge growth demands, its pretty damn hard to find somebody out there with a better story –

  • They have the biggest ecosystem in the NAS space (3rd parties writing to their API’s like crazy)
  • They have some of the tightest integration with the higher elements of the stack
  • They have a great story for VDI using their PAM cards and ASIS de-dupe technology

And that’s not a comprehensive list.  So even if you find somebody who has abetter scale-out story, you can almost bet they don’t do all the good stuff NetApp do.

Also, in the SMB space the fact that their boxes do multi-protocol (CIFS, NFS, iSCSI, FC, FCoE) can be a great benefit.  And you generally don’t get the unmanageable sprawl in the SMB space either.

Conclusion

Yes NetApp technology and architecture is being pushed and there is sprawl out there.  Yes the Spinnaker integration is taking an age.  But you’d be hard pressed to find anybody with a better NAS and unified story.

Obviously nobody is truly unbiased or independent, but that’s my penny’s worth.

11 thoughts on “NetApp: Some good and some errrrr…..

  1. Pingback: Tweets that mention NetApp: Some good and some errrrr….. – Technical Deep Dive -- Topsy.com

  2. Nigel Poulton Post author

    No probs Alex, it’s my honest opinion.

    Now stop tweeting and get cracking on improving your scale-out 😉

  3. Radek

    Nigel,

    As the person responsible for teaming up you four mustakteers (no wait, they're french!) I can only say had Chris posted an even-handed blog like this, none of the resulting commentary would have happened 🙂

    Radek

  4. Paul Stringfellow

    good piece this and pretty balanced view…Storage sprawl (along with server sprawl and App silo's) does seem to be a bit of an enterprise issue anyway…due mainly to how they tend to operate, grow by aquisition etc…

    hence the real hunger to move IT to a more "dyanmic" approach, with correct pools of resource been built, with virtual servers, apps, centralised desktop etc…and of course storage… that maybe we'll see this sprawl reduced…

    and to coin a phrase by 1,000,000% unbiased opinion *cough* NetApp seem to be better placed to be the underpinning storage provider for that infrastructure, with dedupe, thin provisioning, cloning, PAM cards…all of these things providing the optimum in "dynamic" storage…

    nice article though…

  5. Nigel Poulton Post author

    Radek,

    I can see that on one hand it may appear that Chris was dabbling in “tabloid style blogging” as I think you put it. However, good healthy debates like what Chris has generated are good for the vendors. There’s nothing worse for vendors than customers who think the sun shines out of their vendors’ rear ends. Vendors need to be challenged (constructively) and kept on their toes. In fact I spend most of my working life pressing my vendors 😉

    Paul,

    Thanks for your comments. I agree that NetApp are well placed for many use cases. However, they need to work on their fit for large scale deployments as choosing to deploy NetApp at scale is pretty much opening up the doors to your data centre and inviting the local rat population in to make a large ugly nest 😉

    Nigel

  6. Chris M Evans

    So am I Chevvy Chase or Steve Martin?  I assume Storagebod is Martin Short, purely because of his aserbic wit. 

     

    Chris

  7. Pingback: The Storage Architect » Blog Archive » Mad, Bad and Dangerous to Know

  8. visiotech

    Most Unified Storage products in the industry are "good enough" with one type of data transfer but rarely with all three protocols at once.  Also if you have a true Enterprise storage need you rarely place all your eggs in one basket.  You have more than one vendor and more often more than a dozen of storage products devices. 

     

    Today no one show their performance metrics.  They all show their added software widgets.  Even if servers are nearly 50x faster than disks.  Anyone using SSD can see where the bottleneck is…disk need to improve by a major factor.

     

    If storage grow that fast is because no one care about it.  When no disk supplier can deliver bigger disks (soon), they will start manage data.  Otherwise it is just a big dump that no one care about.

  9. Pingback: Tweets that mention NetApp: Some good and some errrrr….. – Technical Deep Dive -- Topsy.com

  10. Pingback: NetApp cluster functionality vs competitors’ clustered file systems | Datacentre Management . org

Leave a Reply

Your email address will not be published. Required fields are marked *


*

You can add images to your comment by clicking here.