Message-ID: <44E308B2.8030008@pbnet.dk>
Date: Wed, 16 Aug 2006 13:59:46 +0200
From: =?ISO-8859-1?Q?Poul_M=F8ller_Hansen?= <freebsd@pbnet.dk>
User-Agent: Thunderbird 1.5.0.5 (Windows/20060719)
MIME-Version: 1.0
To: pgsql-general@postgresql.org
Subject: Segmentation Fault
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Last night one of my databases broke down temporary because of a 
segmentation fault.
It has only happended this time and the database was fully recovered 
afterwards,
but I was wondering what I can do anything to prevent it from happening 
again

It happened while the backup was running (pg_dump & pgdumpall)
Here are some details from the logs etc.

The system is running Ubuntu Linux and I'm using the PostgreSQL package 
from the dapper repository:
uname -a
Linux db 2.6.15-26-amd64-server #1 SMP Fri Jul 7 20:02:26 UTC 2006 
x86_64 GNU/Linux

select version()
PostgreSQL 8.1.4 on x86_64-pc-linux-gnu, compiled by GCC gcc-4.0.gcc-opt 
(GCC) 4.0.3 (Ubuntu 4.0.3-1ubuntu5)

pgsql log
2006-08-16 00:38:22 CEST - LOG:  server process (PID 4792) was 
terminated by sig
nal 11
2006-08-16 00:38:22 CEST - LOG:  terminating any other active server 
processes
2006-08-16 00:38:22 CEST - WARNING:  terminating connection because of 
crash of
another server process
2006-08-16 00:38:22 CEST - DETAIL:  The postmaster has commanded this 
server pro
cess to roll back the current transaction and exit, because another 
server proce
ss exited abnormally and possibly corrupted shared memory.
2006-08-16 00:38:22 CEST - HINT:  In a moment you should be able to 
reconnect to
 the database and repeat your command.

DETAIL and HINT repeated for every connection

2006-08-16 00:38:23 CEST - LOG:  all server processes terminated; 
reinitializing
2006-08-16 00:38:23 CEST - LOG:  database system was interrupted at 
2006-08-16 0
0:36:21 CEST
2006-08-16 00:38:23 CEST - LOG:  checkpoint record is at 5/4F9FDC00
2006-08-16 00:38:23 CEST - LOG:  redo record is at 5/4F9B3558; undo 
record is at
 0/0; shutdown FALSE
2006-08-16 00:38:23 CEST - LOG:  next transaction ID: 5408607; next OID: 
30199
2006-08-16 00:38:23 CEST - LOG:  next MultiXactId: 1; next 
MultiXactOffset: 0
2006-08-16 00:38:23 CEST - LOG:  database system was not properly shut 
down; aut
omatic recovery in progress
2006-08-16 00:38:23 CEST - FATAL:  the database system is starting up
2006-08-16 00:38:23 CEST - LOG:  redo starts at 5/4F9B3558
2006-08-16 00:38:23 CEST - LOG:  record with zero length at 5/4FB63C18
2006-08-16 00:38:23 CEST - LOG:  redo done at 5/4FB63BE8
2006-08-16 00:38:26 CEST - LOG:  database system is ready
2006-08-16 00:38:26 CEST - LOG:  transaction ID wrap limit is 
1073864149, limite
d by database "db"

At 00:36:21 this was happening in the pgsql log
2006-08-16 00:36:21 CEST - LOG:  duration: 14673.110 ms  statement: 
EXECUTE <unn
amed>  [PREPARE:  select * from app.insert_unitstat($1,$2,$3,$4,$5,$6,$7,
$8,$9,$10,$11,$12,$13,$14,$15,$16,$17,$18,$19,$20,$21)  as result]
2006-08-16 00:36:21 CEST - LOG:  duration: 8730.029 ms  statement: 
EXECUTE <unna
med>  [PREPARE:  select * from app.insert_unitstat($1,$2,$3,$4,$5,$6,$7,$
8,$9,$10,$11,$12,$13,$14,$15,$16,$17,$18,$19,$20,$21)  as result]
2006-08-16 00:36:21 CEST - LOG:  duration: 5982.330 ms  statement: 
EXECUTE <unna
med>  [PREPARE:  select * from app.insert_unitstat($1,$2,$3,$4,$5,$6,$7,$
8,$9,$10,$11,$12,$13,$14,$15,$16,$17,$18,$19,$20,$21)  as result]
2006-08-16 00:36:21 CEST - LOG:  duration: 10404.601 ms  statement: 
EXECUTE <unn
amed>  [PREPARE:  select * from app.insert_unitstat($1,$2,$3,$4,$5,$6,$7,
$8,$9,$10,$11,$12,$13,$14,$15,$16,$17,$18,$19,$20,$21)  as result]

These statements are called in a plpgsql function and the function is 
called via JDBC
using postgresql-8.1-407.jdbc3.jar

dmesg
[2425253.737383] postmaster[4792]: segfault at 00002aaab6f0e000 rip 
00002aaaab73795b rsp 00007fffff8c9228 error 4


Any suggestions ?

Thanks,
 Poul