MemSQL In-Memory Database Combines Column and Row Stores

The rise of in-memory computing has set the stage for the convergence of transactional and analytics applications within the same platform. MemSQL is taking organizations a step closer to achieving that goal with the release today of version 3.0 of its namesake in-memory database that now supports column store in addition to traditional relational data stores.

MemSQL CEO Eric Frenkiel says that as the line between transaction processing and analytics applications continues to blur, it’s clear that organizations want to be able to create applications that leverage analytics in real time to inform how any given set of transactions should be processed. MemSQL now allows that to occur by enabling transaction processing and analytics applications to be processed in real time within the same in-memory database.

To allow that to happen smoothly, Frenkiel says MemSQL 3.0 comes with built-in utilities for transferring data between formats that eliminate the need for costly and time-consuming extract, transform and load (ETL) tools.

Fresh off raising an additional $35 million in funding, MemSQL is part of a new class of databases that run completely in-memory. Frenkiel says MemSQL is different from other approaches in that it is now fully capable of running both transaction and analytics application workloads spanning hundreds of terabytes of data running on standard x86 servers using standard SQL.

In-memory computing is clearly the next big thing in enterprise IT. The challenge facing IT organizations now is figuring out which application workloads to move into memory first, given the relative tradeoffs associated with acquiring that capability versus continuing to run applications on existing legacy infrastructure.