> Hello! > > I'm running 9.1.11 on CentOS. I'm getting several messages like this in my logs: > 2014-09-12 01:44:25.583 PDT,,,30540,,4dbffb0c.7b5b,5,,2014-09-12 01:41:42 PDT,74/868,0,LOG,55P03,"skipping analyze of ""foo"" --- lock not available",,,,,, > > Unfortunately, I have several databases with tables named 'foo'. I was expecting a db name in the log message; can someone explain why it's not there?
Fix your log_line_prefix setting to include DB name and then HUP your server.
humm... This looks like CVS log, so the database name would be there if it was a backend process, independent of log_line_prefix. Although looks like this message came from autovacuum process, and so it won't give you the database name.
Perhaps it should have qualified as dbname.schemaname.tablename just like "automatic vacuum" and "automatic analyze" messages do. One could do a feature request for that.
I Suspect that your log_line_prefix was not correctly configured in postgresql.conf.
Try to configure log_line_prefix = '%t [%p]: [%l-1] user=%u db=%d' I think this may help you.