Join Newsletter
Trusted Business Advisors, Expert Technology Analysts

Taneja Blog

Taneja Blog / Data Center Systems / Software Defined/Virtualized Infrastructure

Compressing Data for Performance: Pernix’s Latest Release Squeezes Into RAM

As machines are available with ever more memory in them, we've been seeing that memory put to a lot of good uses lately. Today Pernix Data released FVP 2.5 which updates their big 2.0 release that brought memory into their server-side storage acceleration solution along with the use of flash. Imagine if you could pool server memory across the virtual cluster and use it for a very fast, and protected persistent storage tier. That's pretty much what Pernix FVP does with their Distributed Fault Tolerant Memory (DFTM) design.

We really like Pernix for how it can drop in to existing infrastructure as a software load. You can easily pool flash and RAM for use as high performance storage tier, automatically accelerating whatever storage you are using underneath. Pernix likes to present this as "decoupling" storage performance from capacity, freeing up capacity expansion decisions to made orthogonally from addressing performance needs.

With FVP 2.5, Pernix has now added in-memory compression to their DFTM (now DFTM-Z). This compression is async so it doesn't impact performance. And we are told fitting more data in memory helps extend the Pernix performance storage tier to cover more if not all workloads.  Thus in this case compression serves as a storage acceleration technique rather than as a simple capacity multiplier.

There are a couple of other interesting things here including an intelligent IO profiling feature to figure out which IO should be served by FVP and which should remain direct to storage.  If you've wanted to squeeze more out of your existing virtual server flash/RAM resources, you might want to check out this latest Pernix release.

Bookmark and Share
  • Premiered: 02/25/15
  • Author: Mike Matchett
Topic(s): Pernix Data Decoupled Architecture Storage Acceleration RAM Flash


There are no comments to display. Scroll down to leave your own!


Leave a Comment

You must be logged in to comment. Click here to log in or register if you don't have an account.