Autovacuum seems to block database: WARNING worker took too long to start - Mailing list pgsql-admin

Hi all,

I'm having a strange problem with postgres & autovacuum

Everything is working fine until I start getting the following errors ... and postgres stops working shortly after (it stops accepting connections)

2010-11-13 12:34:08.599 CET|1|||7104||4cde77b0.1bc0|2010-11-13 12:34:08 CET|1/44303|0|| LOG:  automatic vacuum of table "mrs.pg_catalog.pg_statistic": index scans: 1

        pages: 0 removed, 189 remain

        tuples: 132 removed, 4587 remain

        system usage: CPU 0.00s/0.00u sec elapsed 0.04 sec

2010-11-13 13:24:40.998 CET|2|||3300||4cdc2ae6.ce4|2010-11-11 18:41:58 CET||0|| WARNING:  worker took too long to start; cancelled

2010-11-13 13:25:41.126 CET|3|||3300||4cdc2ae6.ce4|2010-11-11 18:41:58 CET||0|| WARNING:  worker took too long to start; cancelled

2010-11-13 13:26:41.254 CET|4|||3300||4cdc2ae6.ce4|2010-11-11 18:41:58 CET||0|| WARNING:  worker took too long to start; cancelled

.........

.........

infinite errors like this ...

This situation obviously creates serious problems for us!! He have to restart the database to make it work again.

Some info of our system

OS: Centos 5.5

Kernel: 2.6.18-194.26.1.el5

Postgres version: 8.4.5 (installation out-of-the-box using yum)

Server memory: 28GB

Autovacuum configuration

autovacuum = on

vacuum_cost_delay = 10

log_autovacuum_min_duration =   0

autovacuum_naptime = 60

autovacuum_vacuum_threshold = 50

autovacuum_vacuum_cost_delay = 20ms

Shared memory value

shared_buffers = 917504

I've read the problem seems to be autovacuum using all resources and leaving the system unasable (not allowing to accept more connections), but I've not read the way to fix it.

Does upgrading to postgres 9.0 help?

Any help would be appreciated!

Thanks

Pablo