Stuart Miniman, a senior analyst at Marlborough, Mass.-based research firm Wikibon, recently wrote a report entitled...
"Server SAN Market Definition". He defined the server SAN as "a pooled storage resource comprising more than one storage device directly attached to separate multiple servers," where the communication between the direct-attached storage occurs via a high-speed interconnect. SearchStorage senior writer Carol Sliwa asked Miniman to predict how server SAN adoption will play out.
Do you foresee IT organizations blending server SANs with their existing storage infrastructure, or do you foresee server SANs overtaking all other forms of storage in the future?
One of the nice things about a server SAN is [that] it should be able to move in on a project basis. If you have an application that you want to roll this into, and you either need to refresh or buy new infrastructure, you might be able to buy this architecture, and this would handle both your compute and storage. As you add use cases and bring workloads into the server SAN, it can grow and really allow you to migrate your existing environments over, because, as we know, most customers aren't going to [rip and replace them].
Where I would say that there is faster adoption of this type of architecture is with service providers who are much more likely to roll out a very large application and need something that is highly scalable and low cost while still having good performance. These architectures are ideal for them. One of the things I found interesting is that when ScaleIO, [which EMC bought], first came out of stealth and went to market, some of their first targets were service providers that were looking to compete with Amazon Elastic Block [Store] service. So, you're talking about thousands of nodes, not just a typical data center that might be dozens or hundreds of nodes.
Related Q&A from Carol Sliwa
Wikibon CTO David Floyer said the greatest challenge with an in-memory database from a storage perspective is recovery in the event of a problem. Continue Reading
Wikibon CTO David Floyer provides advice on how storage IT pros can support the recovery of an in-memory database to avoid hours of downtime. Continue Reading
OpenStack Swift uses regions, zones, servers, drives for data placement and places three copies of each object in unique-as-possible locations. Continue Reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.