hdd1

Great Post on Disk Drive Futures from Hitachi

By On 4 October, 2011 · 9 Comments · In Storage

If you haven’t read it already, I urge you to have a look at the following post on Claus Mikkelsen’s (@yoclaus) blog over at HDS.  It’s a guest post from Ian Vogelesang, who seems to have held many jobs in HGST and Hitachi, but focusing on the disk drive industry.  I warn you it’s highly technical and a long post but it contains some great insights into HDD technology and how Hitachi arrays work with SATA drives.  More of this please Claus & Co!

Here’s the URL again:

http://blogs.hds.com/claus/2011/09/putting-hdd-product-trends-into-perspective-a-subsystem-view.html

Tagged with: • • • •
 
If you enjoyed this article, please consider sharing it!
Reddit Facebook Twitter Delicious Mixx StumbleUpon Digg

  • Pingback: Great Post on Disk Drive Futures from Hitachi | Data Storage

  • http://www.hitachi-storage.com Erwin van Londen

    Ian is a great guy. (He’s Dutch you know. really :-) )

    I had him on my first HDS Internal PODcast called Technocast. Its awesome to hear all the technicalities that come into play on that stuff. Really amazing.

    Cheers,
    Erwin

  • Pingback: Great Post on Disk Drive Futures from Hitachi | The Storage Architect

  • Pingback: Great Post on Disk Drive Futures from Hitachi | The Storage Architect

  • Pingback: Great Post on Disk Drive Futures from Hitachi | Storage news | Scoop.it

  • Kamran Sadique

    Lots of useful information about the co-relation between drives and subsystems…very insightful.

    I do have a question though, how does XIV performs better with SATA drives in comparison to AMS2xxx series?

    • admin

      Kamran

      XIV architecture is different to that of a standard modular array. First, it consists of multiple nodes, not just two. Second it uses RAID-1 for all data. A single block of data is written to one node and mirrored to another node to create a block-based RAID pair. All nodes and all disks participate in handling primary and mirrored I/O all of the time for all LUNs/volumes. This means every LUN created is spread across every disk. Imagine a single SATA drive can do at max 100IOPS. With 180 in an array, that gives you 18,000IOPS capability. Compare that to a RAID group of 8 drives in an AMS2X with 6+2 configuration, you’d get out at best 800IOPS without even looking at the RAID overhead. This is how XIV delivers better apparent performance.

      Like everything, there’s no free lunch. Reading and writing from all disks means a double disk failure (a second disk failing in the array before the data from the first failure has been recreated) can wipe out significant amounts of your data – on EVERY volume. A double disk failure can therefore be catastrophic. Second, XIV can’t offer any QOS. Everything is RAID-1. Hardly power efficient in these times of green computing. There’s also no ability to segregate LUNs or group them to keep some applications separate for better performance. Every host is treated the same.

      Like all architectures, if you understand XIV and are happy with the risks, then it’s worth looking at the cost and see if it works for you.

      Regards
      Chris

      • Kamran Sadique

        Thanks, Chris. Good pointers….and I was actually comparing it to do HDP Pool wide striping…

        but of course, no vendor is perfect and they all have their priorities when it comes to design…

        cool site, btw.

  • http://computersupportowl.com/2012/computer-support-this-ought-to-help/ theblogtroll

    do they have solid state drives?

    Greg
    Computer Support

Subscribe to the Blog!

Subscribe to the Blog!

Don’t forget to subscribe to the blog for more content and monthly emails!