Re: initdb - segmentation fault - Mailing list pgsql-general

From Tom Lane
Subject Re: initdb - segmentation fault
Date
Msg-id 7720.1013447911@sss.pgh.pa.us
Whole thread Raw
In response to Re: initdb - segmentation fault  (Bradley Brown <bradley@segrestfarms.com>)
List pgsql-general
Bradley Brown <bradley@segrestfarms.com> writes:
>     I am including the backtrace form the gdb on the core produced my
> initdb.

Drat.  I was hoping for some symbolic information --- but it looks like
the RPMs you are using have been stripped of all debug symbols.

One possibility, though I never before heard of it causing a crash in
initdb, is that there's a known bug in strcoll() in some versions of
glibc, which does cause crashes during VACUUM ANALYZE if the right data
pattern crops up.  I think that a never-updated copy of RHL 7.1 might
have the buggy version of strcoll().  Try updating to the latest glibc
available for RHL 7.1 and see if it helps.

            regards, tom lane

pgsql-general by date:

Previous
From: Bradley Brown
Date:
Subject: Re: initdb - segmentation fault
Next
From: Doug McNaught
Date:
Subject: Re: getting currval from a sequence - BUG ?