Sorry, I wasn't clear.
Autovacuum runs with default setting, I believe it's daily, or whenever it feels like.
When autovacuum was disabled, we had this problem once every 2 months.
With autovacuum enabled, we had this problem once in every six month. It seems autovacuum could only delay the event, but not prevent it.
On Fri, Aug 10, 2012 at 3:49 PM, François Beausoleil
<francois@teksol.info> wrote:
Le 2012-08-10 à 07:53, Laszlo Fogas a écrit :
> We are running Postgres 8.3 on our production servers on Amazon EC2.
>
> We have a reoccurring problem of slowness initially in every 2 months, after enabling autovacuum every 6 months what only full vacuum can solve. It's kind of a problem as it requires 2hrs downtime and we want to avoid that. What we are doing now is moving to Postgres 9.1 as a desperate measure, but we would like to understand better the root cause of the problem.
The usual solution is to run autovacuum *more* frequently, not less. It's not perfectly clear, but you say "after enabling autovacuum every 6 months". If that's the case, then it's much too long. autovacuum should be running hourly, if not more often.
http://wiki.postgresql.org/wiki/Tuning_Your_PostgreSQL_Server#autovacuum_max_fsm_pages.2C_max_fsm_relations talks about running more frequently.
Hope that helps!
François Beausoleil
