Ask the Expert

Key steps in purchasing/implementing a shared data storage subsystem

We're a small organisation that houses our data in direct-attached storage (DAS) and network-attached storage (NAS) boxes. We need an up-to-date data storage infrastructure to cope with several terabytes of unstructured and database information, but we've never been through the process of procuring a storage system as such. What are the key steps in purchasing and implementing a shared data storage subsystem?

    Requires Free Membership to View

Here are some of the steps I'd recommend when implementing a shared data storage subsystem:

  • Determine how much data is already housed in your current infrastructure.
  • Add to the figure above an estimate for the amount of data you might accrue over the next five years, and then double that number.
  • Pick an iSCSI solution or some DAS hardware for a Sun Microsystems Open Storage system to address the capacity estimate from above and to provide room for growth. If you decide to use plain iSCSI, build a volume manager to share the data.

You should introduce a document management system (DMS) at the same time, or else your unstructured data will likely lose most of its value. You should create a volume with plenty of headroom for your database data, a volume with at least a year of headroom to migrate your legacy data to, and you should create a final volume to store the data for your DMS.

This was first published in November 2009

 

COMMENTS powered by Disqus  //  Commenting policy