Hi Mary, On 2), Neal said you could reproduce the problem without fossology. So what script did you use?
Thanks, Bob On Dec 12, 2008, at 9:45 AM, Laser, Mary wrote: > Hi bob, > redirect_stderr = on # Enable capturing of stderr > into log > log_directory = 'pg_log' # enabled for 2 minute > rotation test > log_filename = 'postgresql-%Y-%m-%d_%H%M%S.log' # enabled for 2 > minute rotation test > log_rotation_age = 2 # changing to every 2 minutes to try > and hang scheduler. > > WRT 2) below, I'm not sure I'm following you. But, what we did was > reloaded a page in the UI that generated a query. Sometimes it > would come back with the info we were requesting (query was carried > out on a non-hung connection) and sometimes not (query was trying to > use a hung connection). Eventually the hung apache connection would > log us out. > > Mary > >> -----Original Message----- >> From: Gobeille, Robert >> Sent: Friday, December 12, 2008 9:34 AM >> To: Laser, Mary >> Cc: Krawetz, Neal; [email protected] >> Subject: Re: [FOSSology] Hang identified >> >> Could I have some more details (or just confirm that the following is >> correct): >> >> 1) set log_rotation_age = 2 in postgresql.conf What do you >> set the other log_rotation variables to? >> >> 2) write a php script that opens a connection and loops on a >> query Report all errors returned by the query >> >> Result is that in ~2minute, a query will issue to the db, and >> will never return. >> >> Do I have this right? >> >> Thanks, >> Bob >> >> >> >> On Dec 12, 2008, at 8:57 AM, Laser, Mary wrote: >> >>> Yes. This is how I was able to reproduce the problem. By >> forcing the >>> log to rotate every 2 minutes, we were able to cause a >> scheduler hang >>> AND we observed apache db connections hanging, too. >>> >>> Mary >>> _______________________________________________ fossology mailing list [email protected] http://fossology.org/mailman/listinfo/fossology
