Mark Lewis <mark(dot)lewis(at)mir3(dot)com> writes:
> But he's using 8.1.4-- in that version, an explain analyze would list
> the time taken to go through triggers, so the fact that we don't see any
> of those lines means that it can't be constraint checking, so wouldn't
> it have to be the index update overhead?
Well, it can't be foreign key checking. Could have been an expensive
function in a CHECK constraint, maybe...
regards, tom lane
In response to
pgsql-performance by date
|Next:||From: James Cloos||Date: 2006-12-13 18:48:10|
|Subject: Optimizing a query|
|Previous:||From: Tim Jones||Date: 2006-12-13 18:42:20|
|Subject: strange query behavior|