Hi,I have this server that I use as db database. It's decent box Ubuntu, 2GB, 
AMD Barton 2.8Gb L2 2Mb. DB version is 7.4.7 - that version was the only one 
available at that time. I have it for about 2 years in similar configuration. 
Lately I've notices that a pack of postmaster(4-22) processes ran by postgres 
user are taking over almost all the CPU. If I restart or reboot it calms down 
but picks up again after a while. I do use psqladmin to vaccum the database but 
I have nothing schedule to run. Data is added on a daily basis and is under 100 
records a day.My plan is to enable all posible logging and take it from there. 
I checked syslog there is no error there, I know that too many files open can 
be a problem. I see none of those or similar. Also I don't see any problem into 
the syslog. I run a web app that has a connection pool of 25 connection and 
what appears to happen is that those connection are not finishing or they are 
very, very slow. Too me it looks that they are waiting on something. Is there a 
way to see what each postmaster runs or do ?What's the best approach in this 
case ? Anyone dealt with similar cases, what was the culprit ?Thanks,MC

Reply via email to