Thomas Bamesberger <[EMAIL PROTECTED]> writes:
> My PostgreSQL Database has been up and running exactly for one year, but
> last week after rebooting the error message "fixrdesc: no pg_class entry
> for pg_class" appeared on the screen. So is there any way to solve this
> problem or to recover the data at least?
Ouch. You have a problem :-(
Are you seeing this failure when you connect to any database in the
cluster, or just one? If the former, I'd guess that you've been
ignoring messages about needing a cluster-wide VACUUM for a little too
long, and you're now hosed due to transaction wraparound. If the
latter, it's probably localized data corruption in that database's
system catalog. Which is not necessarily any better news, but there
are a couple of desperation measures you can try:
1. Remove the "pg_internal.init" file in the broken database, then see
if you can start a backend.
2. Try to start a standalone backend and reindex the system catalogs.
(See the REINDEX man page for details on this rather arcane procedure.)
regards, tom lane
---------------------------(end of broadcast)---------------------------
TIP 5: Have you checked our extensive FAQ?
http://www.postgresql.org/docs/faqs/FAQ.html
