If you haven’t heard about Hekaton by now, you should think about reading up on it. There’s some information from TechNet, a keynote video, and more. Despite the ranting from Oracle’s Bob Evans in 2012 that this was “vaporware”, it’s coming in SQL Server 2014. The latest demos don’t show 100x increase in query speed, but they 30x or so numbers I’ve seen are very impressive.
Hekaton is an in-memory database technology, actually called In-Memory OLTP in Books Online. This isn’t a re-invention of the pintable concept from SQL Server 2000. Instead it’s a very well though out architecture and rather interesting in the way it’s implemented. The keynote video gives you a deep overview, though to use the tables, most of us don’t need to know the internals of how they work.
However we do need to look at the restrictions, err requirements. Since these tables will exist in memory, obviously you need more memory, but more than you thought. The recommendation is twice the expected table size, to account for versioning. In addition, this memory is in addition to the buffer pool and other memory your instance already needs. That means much more complex memory calculations and monitoring for DBAs.
The complexity of SQL Server continues to grow over time as more and more features are added to the product. Many of us don’t need much more than the core relational database functionality for many applications, but the in-memory OLTP tables might be one feature that can help dramatically increase the speed of your application.
Video and Audio versions
Today’s podcast features music by Everyday Jones. No relation, but I stumbled on to them and really like the music. Support this great duo at www.everydayjones.com.