Re: Postgres Crashes on Win2K - Mailing list pgsql-hackers

On Thursday 25 May 2006 10:47, chelsea boot wrote:
> Hi
>
>   Can anyone offer advice on this please: Intermittently a postgres process
> appears to crash and the postmaster service needs restarting.  The
> following is entered in the log:
>
>   2006-03-15 09:50:03 LOG:  server process (PID 348) was terminated by
> signal 125 2006-03-15 09:50:03 LOG:  terminating any other active server
> processes 2006-03-15 09:50:06 FATAL:  the database system is in recovery
> mode 2006-03-15 09:50:06 LOG:  all server processes terminated;
> reinitializing 2006-03-15 09:50:06 FATAL:  the database system is starting
> up
>   2006-03-15 09:50:06 LOG:  database system was interrupted at 2006-03-15
> 00:16:39 GMT Standard Time 2006-03-15 09:50:06 LOG:  checkpoint record is
> at 0/C1F3E48
>   2006-03-15 09:50:06 LOG:  redo record is at 0/C1F3E48; undo record is at
> 0/0; shutdown FALSE 2006-03-15 09:50:06 LOG:  next transaction ID: 651022;
> next OID: 113846 2006-03-15 09:50:06 LOG:  database system was not properly
> shut down; automatic recovery in progress 2006-03-15 09:50:06 FATAL:  the
> database system is starting up
>   2006-03-15 09:50:06 LOG:  record with zero length at 0/C1F3E88
>   2006-03-15 09:50:06 LOG:  redo is not required
>   2006-03-15 09:50:06 FATAL:  the database system is starting up
>   2006-03-15 09:50:06 LOG:  database system is ready
>
>   This is accompanied by 2 postgres errors in the Application Log of the
> Win2K server: (1) "the application failed to initialize properly
> (0xc0000142). Click on OK to terminate the application" (2) "the execption
> unknown software exception (0xc00000fd) occurred in the application at
> 0x7c59bd01. Click on OK to terminate the application".
>
>   Alternatively I get the same messages as above but with an additional
> WARNING:
>
>   WARNING: terminating connection because of a crash of another server
> process DEATIL: The postmaster has commaned this server process to roll
> back the current transaction and exit, possibly because another server
> process exited abnormally and possibly corrupted shared memory.
>
>   What is the signal 125 and how can I troubleshoot what is causing the
> process to crash as the error occurs at random times eg. 8pm when no-one is
> using the network and no utility is running such as AV or backup or during
> the day?
>
>   I've posted to novice and ports but have been unable to find a solution.
>

Try setting log_error_verbosity to verbose and sending that info in along with 
OS and PG Versions.

-- 
Robert Treat
Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL