back to article Dimension Data cloud goes TITSUP down under... after EMC storage fail

Dimension Data's Australian cloud has been down for over 24 hours after EMC kit failed. The outage is detailed on this status page recording problems with the company's “AU 1” region dating back to the morning of July 2nd. Dimension Data company says the service has since come back in “a degraded state” and that “Services are …

  1. Arachnoid

    Meanwhile something is not right in Gotham city

    July 4, 9:14AM

    During the recovery process we had another incident with the Storage Processor .The Vendor is working to bring them back online. All recovery processes are halted and client servers are offline.

    1. Anonymous Coward
      Anonymous Coward

      Re: Meanwhile something is not right in Gotham city

      http://cloud.dimensiondata.com/au/en/about/resources/datasheet/tiered-storage-data-sheet

      The Dimension Data Tiered storage solution exclusively utilises the EMC VNX Series hardware delivering high-performing unified storage with unsurpassed simplicity and efficiency, optimised for virtual applications running on VMware.

      All tiers of cloud disks are deployed on LUNs powered by a hybrid disk configuration consisting of Nearline SAS disks in a RAID5 configuration. In addition both Standard and High Performance tiers are fronted by an extensive ‘Fast Cache’ solid-state disk infrastructure ensuring the highest levels of transactional performance with increased bandwidth and low latency.

  2. TechYogJosh

    So what?

    These outages are pretty regular in typical enterprise datacenter. However, given its a "cloud story", looks like it is and will get more than its fair share of attention (especially after the Code Space debacle). Incidents like these prevent large clients from leveraging an "off premise" cloud, hosted by a service provider. However, these are the same clients who never adopted "off premise" app hosting either. So I doubt cloud has really changed their thinking from a DR or security perspective. Clients who used external hosting partners earlier are comfortable in leveraging third party cloud providers also (rather than developing cloud on their own assets in their own datacenters). More the things change more they remain the same. Alas !!

  3. Anonymous Coward
    Anonymous Coward

    It aint cloud

    "Dimension Data's Australian cloud" ... ok, some co-lo hosting. Some Windows VM's on a big SAN. You call that cloud.... er no, chortle chortle

  4. Anonymous Coward
    Anonymous Coward

    this is what happens...

    .. when you build something enterprise on 2x9's availability platforms. Yes, I'm looking at you Isilon..

    1. Anonymous Coward
      Anonymous Coward

      Re: this is what happens...

      Anyone repeating the NetApp 2 9s FUD doesn't understand 9s or Isilon ;-)

    2. Anonymous Coward
      Anonymous Coward

      Re: this is what happens...

      Interesting. Isilon has not published any information on 9's. Sounds exactly like the words of a NetApp employee based on very little fact or information. Don't believe the hype.

      There isn't even any Isilon in the Dimension Data cloud offering.

      1. bitpushr

        Re: this is what happens...

        EMC *has* presented about its reliability in the past. Publicly.

        Not everything is a competitive FUD-throwing conspiracy.

        Disclaimer: I am a NetApp employee.

        1. Anonymous Coward
          Anonymous Coward

          Re: this is what happens...

          LOL. Really? Given someone has specifically named a product that doesn't feature in the DiData cloud offerings sounds almost EXACTLY like FUD.

  5. Anonymous Coward
    Anonymous Coward

    "we had another incident with the Storage Processor"

    The use of storage processor would suggest EMC kit, specifically the Clariion/VNX/VNX2 range. Unfortunate for all involved really but there have been quite a few of these cloud services outages over the years and in pretty much all cases hardware failure has been only a small part of the story.

    http://www.channelregister.co.uk/2012/01/16/tieto_vnx5700/

    http://www.itnews.com.au/News/230663,emc-san-at-centre-of-state-meltdown.aspx

    http://searchstorage.techtarget.com/news/1510721/EMC-SAN-failure-blamed-for-Intermedia-hosted-email-outages

  6. This post has been deleted by its author

  7. Anonymous Coward
    Anonymous Coward

    Wot? EMC had a big fail?

    Didn't EMC boast about how robust their gear is? That's why they can have the huge markups.

    Does that mean the boasting is over?

  8. Anonymous Coward
    Anonymous Coward

    Blaim the Storage

    I used to work for HDS when they were blamed for a big outage that was actually down to a procedure error at an outsourcer, I've seen NetApp blamed when it was down to power cuts etc. etc.. Everyone who works with storage knows it's always the storage vendors fault. It's raining, storage's fault, drop iPhone in 'sink', the fault of storage, England out the world cup, actually that might not the down to a storage vendor ... Some time storage does mess up. Best dealing with failures I've seen is Isilon if you set the protection at N+2 or higher (decent size cluster still get 80%+ useable space), but that big file storage (50+TB), not block.

    1. Anonymous Coward
      Anonymous Coward

      Re: Blaim the Storage

      Poor operational procedures are the bane of many storage systems, including, and possibly especially Isilon, but I dont think DD's processes are at fault here, and the problem doesnt sound like Isilon ... they keep mentioning "The storage processor" .. also nobody in their right mind implements VMware on isilon for customer facing production workloads, so this is probably some kind of VNX (I doubt a VMAX would be quite this fragile).

  9. Trevor_Pott Gold badge
    Trollface

    This never would have happened if they'd used proper enterprise storage. Trusting in these newfangled startups just puts customers at risk! That's what I don't use cloud computing.

  10. P. Lee

    The Cloud is a site. Worse, its a site which you don't administer. Plan your DR accordingly.

  11. McVirtual
    Holmes

    Fair's fair

    Anyone with an ounce of common sense wouldn't have a single cloud strategy anyway.

    That's the whole point of X-as-a-Service - you can have "afford/ commodotise" your infrastructure and applications effectively.

    We're already discussing multi-cloud and inter-cloud as essential Enterprise business strategies, so when "things" like this happen to poor ol' DiData and EMC, the business isn't actually affected.

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

Other stories you might like