[PERFORM] Interesting performance behaviour

2004-09-10 Thread Joey Smith
#postgresql on Freenode recommended I post this here. I'm seeing some odd behaviour with LIMIT. The query plans are included here, as are the applicable table and index definitions. All table, index, and query information can be found in a standard dbmail 1.2.6 install, if anyone wants to try

Re: [PERFORM] Interesting performance behaviour

2004-09-10 Thread Joey Smith
shared_buffers = 1000 sort_mem = 1024 effective_cache_size = 1000 effective_cache_size should be much higher, like 3/4 of your available RAM. This is probably the essence of your planner problem; the planner thinks you have no RAM. I set effective_cache_size to 64000 on a machine with

Fwd: [PERFORM] Interesting performance behaviour

2004-09-10 Thread Joey Smith
Accidentally sent directly to Josh. -- Forwarded message -- From: Joey Smith [EMAIL PROTECTED] Date: Fri, 10 Sep 2004 15:57:49 -0600 Subject: Re: [PERFORM] Interesting performance behaviour To: [EMAIL PROTECTED] I see a similar speedup (and change in query plan) using LIMIT 1