News for hackers. Stuff that matters.
Re:Understanding Postgres Performance by cafard on Tuesday October 02, 2012 @04:35PM (#4604107)
The problem you can run into, and this will depend on the state of your optimizer, is that you may have a join, and the presence of the index may give you a nested-loops path when a hash join would be more efficient.
Re:Understanding Postgres Performance by jeltz on Tuesday October 02, 2012 @04:57PM (#4604243)
The PostgreSQL optimizer is cost based and uses statistics to choose join method. So that problem is quite rare and is best debugged using EXPLAIN/EXPLAIN ANALYZE.