Обсуждение: logfiles filling up
I'm running Postgresql 8.1.3 on Windows 2003. logfiles fill with the following line: "SSL SYSCALL error: A blocking operation was interrupted by a call to WSACancelBlockingCall." Any ideas? Thanks.
I will add to my initial post: This is a fresh install with a small database (30MB over 12 tables), no other applications on this server and only one user - it's a development machine. And when I say the logfiles "fill up" I mean the entire partition fills, rapidly, until there is "no room left on device". The last time this happened it took less than 10 minutes for 7GB to fill. Seems to happen every couple of days. Shutting down Postgres and removing the logfiles, and then restarting Postgres fixes things temporarily. Erik Myllymaki wrote: > I'm running Postgresql 8.1.3 on Windows 2003. > > logfiles fill with the following line: > > "SSL SYSCALL error: A blocking operation was interrupted by a call to > WSACancelBlockingCall." > > Any ideas? > > Thanks. > > > ---------------------------(end of broadcast)--------------------------- > TIP 9: In versions below 8.0, the planner will ignore your desire to > choose an index scan if your joining column's datatypes do not > match
Do you actively use SSL on this server? //Magnus > -----Original Message----- > From: pgsql-general-owner@postgresql.org > [mailto:pgsql-general-owner@postgresql.org] On Behalf Of Erik > Myllymaki > Sent: Wednesday, May 03, 2006 3:36 PM > To: pgsql-general@postgresql.org > Subject: Re: [GENERAL] logfiles filling up > > I will add to my initial post: > > This is a fresh install with a small database (30MB over 12 > tables), no other applications on this server and only one > user - it's a development machine. > > And when I say the logfiles "fill up" I mean the entire > partition fills, rapidly, until there is "no room left on device". > > The last time this happened it took less than 10 minutes for > 7GB to fill. > > Seems to happen every couple of days. > > Shutting down Postgres and removing the logfiles, and then > restarting Postgres fixes things temporarily. > > Erik Myllymaki wrote: > > I'm running Postgresql 8.1.3 on Windows 2003. > > > > logfiles fill with the following line: > > > > "SSL SYSCALL error: A blocking operation was interrupted by > a call to > > WSACancelBlockingCall." > > > > Any ideas? > > > > Thanks. > > > > > > ---------------------------(end of > > broadcast)--------------------------- > > TIP 9: In versions below 8.0, the planner will ignore your desire to > > choose an index scan if your joining column's datatypes do not > > match > > ---------------------------(end of > broadcast)--------------------------- > TIP 3: Have you checked our extensive FAQ? > > http://www.postgresql.org/docs/faq >
yes I do. Magnus Hagander wrote: > Do you actively use SSL on this server? > > //Magnus > >> -----Original Message----- >> From: pgsql-general-owner@postgresql.org >> [mailto:pgsql-general-owner@postgresql.org] On Behalf Of Erik >> Myllymaki >> Sent: Wednesday, May 03, 2006 3:36 PM >> To: pgsql-general@postgresql.org >> Subject: Re: [GENERAL] logfiles filling up >> >> I will add to my initial post: >> >> This is a fresh install with a small database (30MB over 12 >> tables), no other applications on this server and only one >> user - it's a development machine. >> >> And when I say the logfiles "fill up" I mean the entire >> partition fills, rapidly, until there is "no room left on device". >> >> The last time this happened it took less than 10 minutes for >> 7GB to fill. >> >> Seems to happen every couple of days. >> >> Shutting down Postgres and removing the logfiles, and then >> restarting Postgres fixes things temporarily. >> >> Erik Myllymaki wrote: >>> I'm running Postgresql 8.1.3 on Windows 2003. >>> >>> logfiles fill with the following line: >>> >>> "SSL SYSCALL error: A blocking operation was interrupted by >> a call to >>> WSACancelBlockingCall." >>> >>> Any ideas? >>> >>> Thanks. >>> >>> >>> ---------------------------(end of >>> broadcast)--------------------------- >>> TIP 9: In versions below 8.0, the planner will ignore your desire to >>> choose an index scan if your joining column's datatypes do not >>> match >> ---------------------------(end of >> broadcast)--------------------------- >> TIP 3: Have you checked our extensive FAQ? >> >> http://www.postgresql.org/docs/faq >>
hm. Ok. The log file contains a PID field, IIRC. Does that PID: 1) Stay the same, or change? 2) Match postmaster, or postgres.exe? If postgres.exe, is it just a backend? 3) If it's just a backend, is it enough to kill off that backend? //Magnus > -----Original Message----- > From: Erik Myllymaki [mailto:erik.myllymaki@aviawest.com] > Sent: Wednesday, May 03, 2006 3:49 PM > To: Magnus Hagander > Cc: pgsql-general@postgresql.org > Subject: Re: [GENERAL] logfiles filling up > > yes I do. > > > Magnus Hagander wrote: > > Do you actively use SSL on this server? > > > > //Magnus > > > >> -----Original Message----- > >> From: pgsql-general-owner@postgresql.org > >> [mailto:pgsql-general-owner@postgresql.org] On Behalf Of Erik > >> Myllymaki > >> Sent: Wednesday, May 03, 2006 3:36 PM > >> To: pgsql-general@postgresql.org > >> Subject: Re: [GENERAL] logfiles filling up > >> > >> I will add to my initial post: > >> > >> This is a fresh install with a small database (30MB over > 12 tables), > >> no other applications on this server and only one user - it's a > >> development machine. > >> > >> And when I say the logfiles "fill up" I mean the entire partition > >> fills, rapidly, until there is "no room left on device". > >> > >> The last time this happened it took less than 10 minutes > for 7GB to > >> fill. > >> > >> Seems to happen every couple of days. > >> > >> Shutting down Postgres and removing the logfiles, and then > restarting > >> Postgres fixes things temporarily. > >> > >> Erik Myllymaki wrote: > >>> I'm running Postgresql 8.1.3 on Windows 2003. > >>> > >>> logfiles fill with the following line: > >>> > >>> "SSL SYSCALL error: A blocking operation was interrupted by > >> a call to > >>> WSACancelBlockingCall." > >>> > >>> Any ideas? > >>> > >>> Thanks. > >>> > >>> > >>> ---------------------------(end of > >>> broadcast)--------------------------- > >>> TIP 9: In versions below 8.0, the planner will ignore > your desire to > >>> choose an index scan if your joining column's > datatypes do not > >>> match > >> ---------------------------(end of > >> broadcast)--------------------------- > >> TIP 3: Have you checked our extensive FAQ? > >> > >> http://www.postgresql.org/docs/faq > >> >
I don't see any kind of PID in the logfiles at all, just lines and lines of: 2006-05-02 23:03:28 LOG: SSL SYSCALL error: A blocking operation was interrupted by a call to WSACancelBlockingCall. 2006-05-02 23:03:28 LOG: SSL SYSCALL error: A blocking operation was interrupted by a call to WSACancelBlockingCall. 2006-05-02 23:03:28 LOG: SSL SYSCALL error: A blocking operation was interrupted by a call to WSACancelBlockingCall. etc. When this trouble happens, these logfiles are each 10-12 MB in size and there are ~10 produced every minute until the disk fills. Magnus Hagander wrote: > hm. Ok. > The log file contains a PID field, IIRC. Does that PID: > 1) Stay the same, or change? > 2) Match postmaster, or postgres.exe? If postgres.exe, is it just a > backend? > 3) If it's just a backend, is it enough to kill off that backend? > > //Magnus > >> -----Original Message----- >> From: Erik Myllymaki [mailto:erik.myllymaki@aviawest.com] >> Sent: Wednesday, May 03, 2006 3:49 PM >> To: Magnus Hagander >> Cc: pgsql-general@postgresql.org >> Subject: Re: [GENERAL] logfiles filling up >> >> yes I do. >> >> >> Magnus Hagander wrote: >>> Do you actively use SSL on this server? >>> >>> //Magnus >>> >>>> -----Original Message----- >>>> From: pgsql-general-owner@postgresql.org >>>> [mailto:pgsql-general-owner@postgresql.org] On Behalf Of Erik >>>> Myllymaki >>>> Sent: Wednesday, May 03, 2006 3:36 PM >>>> To: pgsql-general@postgresql.org >>>> Subject: Re: [GENERAL] logfiles filling up >>>> >>>> I will add to my initial post: >>>> >>>> This is a fresh install with a small database (30MB over >> 12 tables), >>>> no other applications on this server and only one user - it's a >>>> development machine. >>>> >>>> And when I say the logfiles "fill up" I mean the entire partition >>>> fills, rapidly, until there is "no room left on device". >>>> >>>> The last time this happened it took less than 10 minutes >> for 7GB to >>>> fill. >>>> >>>> Seems to happen every couple of days. >>>> >>>> Shutting down Postgres and removing the logfiles, and then >> restarting >>>> Postgres fixes things temporarily. >>>> >>>> Erik Myllymaki wrote: >>>>> I'm running Postgresql 8.1.3 on Windows 2003. >>>>> >>>>> logfiles fill with the following line: >>>>> >>>>> "SSL SYSCALL error: A blocking operation was interrupted by >>>> a call to >>>>> WSACancelBlockingCall." >>>>> >>>>> Any ideas? >>>>> >>>>> Thanks. >>>>> >>>>> >>>>> ---------------------------(end of >>>>> broadcast)--------------------------- >>>>> TIP 9: In versions below 8.0, the planner will ignore >> your desire to >>>>> choose an index scan if your joining column's >> datatypes do not >>>>> match >>>> ---------------------------(end of >>>> broadcast)--------------------------- >>>> TIP 3: Have you checked our extensive FAQ? >>>> >>>> http://www.postgresql.org/docs/faq >>>> > > ---------------------------(end of broadcast)--------------------------- > TIP 5: don't forget to increase your free space map settings
Hmm. I tohught we had PID as default, apparntly we don't. Try setting log_line_prefix to something containing %p (for process pid) - probably "%t %p ", unless you have changed things from the default. Then restart the server and wait for the next time it happens :-( //Magnus > -----Original Message----- > From: Erik Myllymaki [mailto:erik.myllymaki@aviawest.com] > Sent: Wednesday, May 03, 2006 6:17 PM > To: Magnus Hagander > Cc: pgsql-general@postgresql.org > Subject: Re: [GENERAL] logfiles filling up > > I don't see any kind of PID in the logfiles at all, just > lines and lines of: > > 2006-05-02 23:03:28 LOG: SSL SYSCALL error: A blocking > operation was interrupted by a call to WSACancelBlockingCall. > > > 2006-05-02 23:03:28 LOG: SSL SYSCALL error: A blocking > operation was interrupted by a call to WSACancelBlockingCall. > > > 2006-05-02 23:03:28 LOG: SSL SYSCALL error: A blocking > operation was interrupted by a call to WSACancelBlockingCall. > > etc. > > When this trouble happens, these logfiles are each 10-12 MB > in size and there are ~10 produced every minute until the disk fills. > > Magnus Hagander wrote: > > hm. Ok. > > The log file contains a PID field, IIRC. Does that PID: > > 1) Stay the same, or change? > > 2) Match postmaster, or postgres.exe? If postgres.exe, is it just a > > backend? > > 3) If it's just a backend, is it enough to kill off that backend? > > > > //Magnus > > > >> -----Original Message----- > >> From: Erik Myllymaki [mailto:erik.myllymaki@aviawest.com] > >> Sent: Wednesday, May 03, 2006 3:49 PM > >> To: Magnus Hagander > >> Cc: pgsql-general@postgresql.org > >> Subject: Re: [GENERAL] logfiles filling up > >> > >> yes I do. > >> > >> > >> Magnus Hagander wrote: > >>> Do you actively use SSL on this server? > >>> > >>> //Magnus > >>> > >>>> -----Original Message----- > >>>> From: pgsql-general-owner@postgresql.org > >>>> [mailto:pgsql-general-owner@postgresql.org] On Behalf Of Erik > >>>> Myllymaki > >>>> Sent: Wednesday, May 03, 2006 3:36 PM > >>>> To: pgsql-general@postgresql.org > >>>> Subject: Re: [GENERAL] logfiles filling up > >>>> > >>>> I will add to my initial post: > >>>> > >>>> This is a fresh install with a small database (30MB over > >> 12 tables), > >>>> no other applications on this server and only one user - it's a > >>>> development machine. > >>>> > >>>> And when I say the logfiles "fill up" I mean the entire > partition > >>>> fills, rapidly, until there is "no room left on device". > >>>> > >>>> The last time this happened it took less than 10 minutes > >> for 7GB to > >>>> fill. > >>>> > >>>> Seems to happen every couple of days. > >>>> > >>>> Shutting down Postgres and removing the logfiles, and then > >> restarting > >>>> Postgres fixes things temporarily. > >>>> > >>>> Erik Myllymaki wrote: > >>>>> I'm running Postgresql 8.1.3 on Windows 2003. > >>>>> > >>>>> logfiles fill with the following line: > >>>>> > >>>>> "SSL SYSCALL error: A blocking operation was interrupted by > >>>> a call to > >>>>> WSACancelBlockingCall." > >>>>> > >>>>> Any ideas? > >>>>> > >>>>> Thanks. > >>>>> > >>>>> > >>>>> ---------------------------(end of > >>>>> broadcast)--------------------------- > >>>>> TIP 9: In versions below 8.0, the planner will ignore > >> your desire to > >>>>> choose an index scan if your joining column's > >> datatypes do not > >>>>> match > >>>> ---------------------------(end of > >>>> broadcast)--------------------------- > >>>> TIP 3: Have you checked our extensive FAQ? > >>>> > >>>> http://www.postgresql.org/docs/faq > >>>> > > > > ---------------------------(end of > > broadcast)--------------------------- > > TIP 5: don't forget to increase your free space map settings >