Ivano Luberti wrote:
This is what I was thinking , but today I was able to look at the processes running while a client was doing a bunc of inserts. There was no autovacuum running and every insert was taking many seconds to e executed.
Have you done any basic tuning of the database parameters? From your earlier message, it sounds like you might have checkpoint_segments at its default, which can contribute to these inserts taking so long. See http://wiki.postgresql.org/wiki/Tuning_Your_PostgreSQL_Server for a guide to where to start.
Also, if you want to get a better idea what's actually going on with your data, you should be looking at "select * from pg_stat_user_tables" ; that will give you information about things like how many dead rows there are in the table, when autovacuum last did some work, etc. Should be possible to figure out what's different about the use pattern of this table compared to the ones you suggest work as expected by analyzing that data.
P.S. To clean up from one of these messes you might try CLUSTER instead of VACUUM FULL.
-- Greg Smith 2ndQuadrant US Baltimore, MD PostgreSQL Training, Services and Support greg@xxxxxxxxxxxxxxx www.2ndQuadrant.us -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
