On 18 May 2016 at 14:25, Andres Freund <and...@anarazel.de> wrote: > Hi, > > currently we IIRC use linearly sorted datums for the search in > individual btree nodes. Not surprisingly that's often one of the > dominant entries in profiles. We could probably improve upon that by > using an order more optimized for efficient binary search. > > See e.g. http://cglab.ca/~morin/misc/arraylayout/ for benchmarks > showing benefits. >
Some stuff from >10 years ago about cache conscious btree layout as well. That led to adoption of 64kB pages on some benchmarks. I think its a good area of work. -- Simon Riggs http://www.2ndQuadrant.com/ <http://www.2ndquadrant.com/> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
