Re: Performance (was: The New Slashdot Setup (includes MySql server))

Hi,

Hiroshi Inoue:
> What does 'vacuum pg_index' show ?
>
test=> vacuum pg_index;
NOTICE: Skipping "pg_index" --- only table owner can VACUUM it
VACUUM

OK, so I suppose I should do it as the postgres user...
test=> vacuum pg_index;
VACUUM

The debug output says:
DEBUG: --Relation pg_index--
DEBUG: Pages 448: Changed 0, reaped 448, Empty 0, New 0; Tup 34: Vac 21443, Keep/VTL 0/0, Crash 0, UnUsed 0, MinLen 164, MaxLen 164;
Re-using: Free/Avail. Space 3574948/3567176; EndEmpty/Avail. Pages 0/447. CPU 0.46s/0.00u sec.
DEBUG: Index pg_index_indexrelid_index: Pages 86; Tuples 34: Deleted 21443. CPU 0.05s/0.36u sec.
DEBUG: Rel pg_index: Pages: 448 --> 1; Tuple(s) moved: 2. CPU 0.03s/0.03u sec.
DEBUG: Index pg_index_indexrelid_index: Pages 86; Tuples 34: Deleted 2. CPU 0.01s/0.00u sec.

... which helped. A lot.

Thanks, everybody. The first quick benchmark run I did afterwards states
that PostgreSQL is now only half as fast as MySQL, instead of the factor
of 30 seen previously, on the MySQL benchmark test. ;-)

--
Matthias Urlichs | noris network GmbH | smurf(at)noris(dot)de | ICQ: 20193661
The quote was selected randomly. Really. | http://smurf.noris.de/
--
Dorian Graying:
The unwillingness to gracefully allow one's body to show signs
of aging.
-Douglas Coupland, Generation X