In this segment from his recent Storage Decisions presentation, Dennis Martin, president of the analyst firm Demartek LLC, discusses some of the options administrators have when considering how to best use flash as part of an existing storage system.
"You can put it in the storage. Now, by 'in the storage,' I mean big storage arrays; you got those big frames with lots and lots drives in them, from small to big, you can have your scope [of options]," Martin said.
He noted that adding flash to a storage system does not require changes to the server, and implementing flash as cache avoids changes to applications, as well.
"If you don't want to change the server, you just change it in the storage. The storage suddenly gets faster. The server doesn't know any different," said Martin. "If there are caching implementations, again, you don't have to change anything on the server. You can … dedicate part of the system to use as a cache and operate it that way. And, of course, you can share, because you're on the back end, where all the servers and everything can come in."
According to Martin, using flash in a tiered storage system requires some careful consideration.
"If you start doing tiering, where you have fast storage, medium speeds, slower storage, then your apps have to know in some cases that this tiering is going on and the location of the data might be different. It depends on the tiering solution," he said.