Re: Duplicate WHERE condition changes performance and plan

2020-04-15 Thread David Rowley
On Thu, 16 Apr 2020 at 07:56, singh...@gmail.com wrote: > We have an odd issue where specifying the same where clause twice causes PG > to pick a much more efficent plan. We would like to know why. > The EXPLAIN ANALYZE for both queries can be found here:- > Query A: https://explain.depesz.com/s

Using unlogged tables for web sessions

2020-04-15 Thread Stephen Carboni
Hi. Apologies in advance if this is not the right place to ask. I was wondering if anyone was using unlogged tables for website sessions in production. I'm interested if UNLOGGED breaks the prevailing opinion that you don't put sessions in PG.

Duplicate WHERE condition changes performance and plan

2020-04-15 Thread singh...@gmail.com
Hi, We have an odd issue where specifying the same where clause twice causes PG to pick a much more efficent plan. We would like to know why. Query A (this is the 'slow' query): UPDATE problem_instance SET processed = false FROM problem WHERE problem.id = problem_instance.problem_id AND problem.s