6 dec 2007 kl. 20.26 skrev Alvaro Herrera:
Henrik wrote:
I think I have a clue why its so off. We update a value in that
table about
2 - 3 million times per night and as update creates a new row it
becomes
bloated pretty fast. The table hade a size of 765 MB including
indexes and
after vacuum full and reindex it went down to 80kB... I guess I need
routine reindex on this table. Thank god is not big. :)
I suggest you put a lone VACUUM on that table in cron, say once
every 5
minutes, and you should be fine. You shouldn't need a reindex at all.
Instead of cron can't I just have really aggressive autovacuum
settings on this table?
Thanks,
Henke
--
Alvaro Herrera http://www.PlanetPostgreSQL.org/
"Right now the sectors on the hard disk run clockwise, but I heard a
rumor that
you can squeeze 0.2% more throughput by running them counterclockwise.
It's worth the effort. Recommended." (Gerry Pourwelle)
---------------------------(end of
broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to majordomo@xxxxxxxxxxxxxx so that
your
message can get through to the mailing list cleanly
---------------------------(end of broadcast)---------------------------
TIP 5: don't forget to increase your free space map settings
