I can confirm that adding the indexes used by the deferred constraint
triggers solved the issue.
Thank you very much for your suggestions.
On 10/24/07, Giulio Cesare Solaroli <giulio(dot)cesare(at)gmail(dot)com> wrote:
> On 10/24/07, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> > "Giulio Cesare Solaroli" <giulio(dot)cesare(at)gmail(dot)com> writes:
> > > How can I try to isolate the trigger taking so long, in oder to
> > > understand which is/are the missing index(es)?
> > Try SET CONSTRAINTS ALL IMMEDIATE and then EXPLAIN ANALYZE the
> > delete. This should cause all the triggers to run within the
> > scope of the EXPLAIN ANALYZE, and you'll be able to see which
> > one(s) are slow. (This assumes you're running a recent release
> > of PG; I think EXPLAIN shows trigger times since 8.1 or so.)
> I was thinking about something similar after writing the last message.
> Thank you very much for your attention!!
> Giulio Cesare
In response to
pgsql-performance by date
|Next:||From: Jean-David Beyer||Date: 2007-10-26 13:04:43|
|Subject: Re: Bunching "transactions"|
|Previous:||From: Gregory Stark||Date: 2007-10-26 08:07:23|
|Subject: Re: [HACKERS] 8.3beta1 testing on Solaris|