Re: full-text search with GiST indexes seems to be getting slower since 9.1

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: full-text search with GiST indexes seems to be getting slower since 9.1
Date: 2020-05-03 03:09:13
Message-ID: 17387.1588475353@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

Tomas Vondra <tomas(dot)vondra(at)2ndquadrant(dot)com> writes:
> over the past couple of weeks I've been running various benchmarks with
> the intent to demonstrate how the performance evolved since ~8.3. In
> most tests we're doing pretty good, but I've noticed that full-text
> search using GiST indexes is an annoying exception - it's getting slower
> and slower since ~9.1.

Hmmm... given that TS_execute is near the top, I wonder if this has
anything to do with the several times we've mostly-rewritten that
to make it less incorrect for phrase search cases. But that came in
in 9.6 IIRC, and the fixes have only been back-patched that far.
Don't know what might've changed in 9.3.

regards, tom lane

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Noah Misch 2020-05-03 06:48:36 Re: implicit declaration of datumIsEqual in parse_coerce.c
Previous Message Justin Pryzby 2020-05-03 02:42:15 Re: pg_ls_tmpdir to show directories and shared filesets (and pg_ls_*)