News for hackers. Stuff that matters.
Re:Sysbench Benchmark for MongoDB – Performance Update by leif on Wednesday June 05, 2013 @10:19AM (#5825706)
The name is a marketing term, you're right. The structure is somewhere in the realm of buffer trees, LSM-trees, $B^\epsilon$ trees, and COLAs. It's hard to give it an academic name precisely because the implementation takes hints from many places, but we hope to publish more soon. B-trees are on the optimal tradeoff curve but there are many other points on that curve, and B-trees heavily favor reads so there's plenty of room to win on the write side. I have a couple of blogs about this actually: http://www.tokutek.com/2011/09/write-optimization-myths-comp... http://www.tokutek.com/2011/10/write-optimization-myths-comp...

There are no comments on this comment yet.