I recently migrated from MSSQL2000 to Postgres 8.3 for Windows, and overall it’s running great.
Using the Process Explorer tool, I've noticed that a child postgres.exe is chewing up 25% of the CPU usage each (we have two dual-core CPUs, presumably it’s chewing up one core). Using SELECT * FROM pg_stat_activity, I located the process id (#3884), and it showed:
datid | datname | procpid | usesysid | usename | current_query | waiting | xact_start | query_start | backend_start | client_addr | client_port |
11511 | postgres | 3884 | 16395 | Brendan.Hill | <IDLE> | f | | 2009-07-28 13:54:50.055+10 | 2009-07-28 13:54:44.508+10 | ### | 3353 |
No processes are waiting. This was taken at 2:19pm, so it had been running for about 25 minutes. This has happened before – we had two runaway processes, and they each chewed up 25% of the CPU for several weeks, before we forcefully shut them down. We had to forcefully shut them down before we could restart the Postgres service.
Other requests are being served, however we’ve had reports of slowness (which sparked off the initial search). I can’t confirm this runaway process is the source of it, but it shouldn’t be happening either way. I’ve killed it just in case.
Any thoughts on what is causing this, or how I could diagnose the problem further?
Regards,
Brendan Hill
Chief Information Officer
Jims Group Pty Ltd
48 Edinburgh Rd
Mooroolbark VIC 3138
www.jims.net
For all Jims IT enquiries: infotech@jims.net
For emergencies: 1300 130 490 (intl +61 4 3456 5776)
