Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-17 Thread hubert depesz lubaczewski
On 12/17/05, Harry Jackson <[EMAIL PROTECTED]> wrote: On 12/17/05, hubert depesz lubaczewski <[EMAIL PROTECTED]> wrote:>  must have missed that one. thanks - we are moving to 8.1 anyway (50-150%> increase in performance for some specific tasks!). For which tasks in particular are you seeing this ki

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-17 Thread Harry Jackson
On 12/17/05, hubert depesz lubaczewski <[EMAIL PROTECTED]> wrote: > must have missed that one. thanks - we are moving to 8.1 anyway (50-150% > increase in performance for some specific tasks!). For which tasks in particular are you seeing this king of gain? Harry -- http://www.hjackson.org http

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-17 Thread hubert depesz lubaczewski
On 12/16/05, Tom Lane <[EMAIL PROTECTED]> wrote: 2005-08-09 17:14  tgl    * src/backend/postmaster/pgstat.c: Make backends that are readingthe pgstats file verify each backend PID against the PGPROC array.Anything in the file that isn't in PGPROC gets rejected as being a    

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-15 Thread Tom Lane
hubert depesz lubaczewski <[EMAIL PROTECTED]> writes: > of course i can make myself a function which will check if pid exists, but > it would generally be better if there was some way to fix the problem in > postgresql itself. Try updating to PG 8.1 ... 2005-08-09 17:14 tgl * src/backen

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-15 Thread hubert depesz lubaczewski
On 12/14/05, Chris Browne <[EMAIL PROTECTED]> wrote: The only answer I have been able to come to is that restarting thepostmaster will clear this all up. this is actually not an option for me. of course i can make myself a function which will check if pid exists, but it would generally be better if

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-15 Thread Gregory S. Williamson
Jim Nasby wrote: > > I too have had issues with "stuck" entries in the pg_stat_activity view -- > > > the real pid is long gone but postgres still thinks the process is there. > > > > It would be real sweet to have a way of cleaning this table with bringing > > > postgres off-line. > > Rather

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-15 Thread Jim C. Nasby
On Wed, Dec 14, 2005 at 02:02:02PM -0800, Gregory S. Williamson wrote: > I too have had issues with "stuck" entries in the pg_stat_activity view -- > the real pid is long gone but postgres still thinks the process is there. > > It would be real sweet to have a way of cleaning this table with brin

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-14 Thread Gregory S. Williamson
obeXplorer LLC -Original Message- From: [EMAIL PROTECTED] on behalf of Chris Browne Sent: Wed 12/14/2005 11:06 AM To: pgsql-general@postgresql.org Cc: Subject: Re: [GENERAL] outdated (bad) information in pg_stat_activity [EMAIL PROTECTED] (hubert depesz lubaczewski) wr

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-14 Thread Chris Browne
[EMAIL PROTECTED] (hubert depesz lubaczewski) writes: > On 12/14/05, Tom Lane <[EMAIL PROTECTED]> wrote: > > > hubert depesz lubaczewski <[EMAIL PROTECTED]> writes: > > my pg_stat_activity is raporting process working on a query for 12 > hours > > straight! (query_start). >

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-14 Thread hubert depesz lubaczewski
On 12/14/05, Tom Lane <[EMAIL PROTECTED]> wrote: hubert depesz lubaczewski <[EMAIL PROTECTED]> writes:> my pg_stat_activity is raporting process working on a query for 12 hours> straight! (query_start).> 2 problems: > 1. this particular query runs typically between 80 and 200 miliseconds!> 2. procp

Re: [GENERAL] outdated (bad) information in pg_stat_activity

2005-12-14 Thread Tom Lane
hubert depesz lubaczewski <[EMAIL PROTECTED]> writes: > my pg_stat_activity is raporting process working on a query for 12 hours > straight! (query_start). > 2 problems: > 1. this particular query runs typically between 80 and 200 miliseconds! > 2. procpid of this backend does not exists anymore in

[GENERAL] outdated (bad) information in pg_stat_activity

2005-12-14 Thread hubert depesz lubaczewski
hi my pg_stat_activity is raporting process working on a query for 12 hours straight! (query_start). 2 problems: 1. this particular query runs typically between 80 and 200 miliseconds! 2. procpid of this backend does not exists anymore in system i found this: http://www.issociate.de/board/post/264