Обсуждение: problem with segmentation fault error

Hello,
I've got problem with PostgreSQL crash (today it was twice).

Usually, average load5 is about 16-20. Today I've run new application and load has increased dramatically.

First time before crash load was fast rising from 30 to 70 (log below)
2010-01-27 10:34:01 CET [3931]: [9-1] host=,user=,db= LOG:  server process (PID 24454) was terminated by signal 11:
Segmentationfault 
2010-01-27 10:34:01 CET [3931]: [10-1] host=,user=,db= LOG:  terminating any other active server processes
2010-01-27 10:34:01 CET [21906]: [1-1] host=X.X.X.X,user=pgmed,db=pgmed WARNING:  terminating connection because of
crashof another server process 
2010-01-27 10:34:01 CET [21906]: [2-1] host=X.X.X.X,user=pgmed,db=pgmed DETAIL:  The postmaster has commanded this
serverprocess to roll back the current transaction and exit, because another server process exited abnormally and
possiblycorrupted shared memory. 

Second time, load was about 70-80 and processor utilization was almost 100% from 1 hour (log below)
2010-01-27 18:46:06 CET [3931]: [16-1] host=,user=,db= LOG:  server process (PID 28357) was terminated by signal 11:
Segmentationfault 
2010-01-27 18:46:06 CET [3931]: [17-1] host=,user=,db= LOG:  terminating any other active server processes
2010-01-27 18:46:06 CET [26377]: [3-1] host=X.X.X.X,user=pgmed_app,db=pgmed WARNING:  terminating connection because of
crashof another server process 
2010-01-27 18:46:06 CET [26377]: [4-1] host=X.X.X.X,user=pgmed_app,db=pgmed DETAIL:  The postmaster has commanded this
serverprocess to roll back the current transaction and exit, because another server process exited abnormally and
possiblycorrupted shared memory. 

I have 8.4.0 version.

Do you have any idea where problem could be? Is it hardware or rather PostgreSQL configuration problem?
I'll be grateful for any help.

Best regards
PG

pg2@onet.eu writes:
> 2010-01-27 10:34:01 CET [3931]: [9-1] host=,user=,db= LOG:  server process (PID 24454) was terminated by signal 11:
Segmentationfault 

> I have 8.4.0 version.

The *first* thing to do is update to 8.4.2 to see if the problem was
already fixed.

If not, you need to try to gather some data.  What query was running in
the crashed backend would be useful to know, and a stack trace would be
even better.

> Do you have any idea where problem could be? Is it hardware or rather PostgreSQL configuration problem?

It could be a hardware problem or a software bug.  How reproducible is
it?

            regards, tom lane