| |||||||||||||
|
| |||||||||||||
Haas: Big ideas [for PostgreSQL]Haas: Big ideas [for PostgreSQL]Posted May 6, 2010 8:23 UTC (Thu) by petereisentraut (subscriber, #59453)In reply to: Haas: Big ideas [for PostgreSQL] by arjan Parent article: Haas: Big ideas [for PostgreSQL]
Indexes on dates are supported just fine. What you are apparently experiencing is an optimizer decision of one plan over another, based on statistics and the optimization cost model. If you think the optimizer is wrong in a particular case, please submit a bug report. But there is no new "feature" here; it already works in theory.
(Log in to post comments)
Haas: Big ideas [for PostgreSQL] Posted May 6, 2010 19:47 UTC (Thu) by jeremiah (subscriber, #1221) [Link]
you might also find that you need to adjust the stats collection a little bit and then rerun vaccum analyze. You'll laso find the a sorted index on your dates works EXTREMELY well.
|
|
||||||||||||