Anyone else seeing coredumps in pgdump? - Mailing list pgsql-hackers

From Tom Lane
Subject Anyone else seeing coredumps in pgdump?
Date
Msg-id 20116.904531574@sss.pgh.pa.us
Whole thread Raw
List pgsql-hackers
I'm having a little trouble updating to the latest cvs tree, because
I can't dump my existing database, neither with the latest pgdump source
nor with that of two weeks ago.  I've run out of steam to debug it
tonight, but it looks like the failure is occuring while pgdump tries
to locate the inheritance parent of a class.

The comments in the immediate vicinity of where the failure is occuring
indicate that pgdump is relying on inheriting classes to have larger
OID numbers than their parents.  Offhand that doesn't seem too
unreasonable, but is it possible that some recent change has broken
this assumption?

Is anyone else seeing similar problems??

            regards, tom lane

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [HACKERS] flock patch breaks things here
Next
From: Bruce Momjian
Date:
Subject: encoding problem