| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | "Barnes, Sandy (Sandra)" <Sandy(dot)Barnes(at)Honeywell(dot)com> |
| Cc: | "'pgsql-hackers(at)postgresql(dot)org'" <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: V7.0 Error: OID ##### no longer exists in pg_databas e |
| Date: | 2001-01-03 20:56:23 |
| Message-ID: | 17847.978555383@sss.pgh.pa.us |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
"Barnes, Sandy (Sandra)" <Sandy(dot)Barnes(at)Honeywell(dot)com> writes:
>> Any ideas on why this error would occur after dropping an existing
>> database and than re-creating it (createdb)?
I assume what you're really talking about is
Database 'xxx', OID nnn, has disappeared from pg_database
?
It's theoretically possible for that to happen if, say, a CREATE
DATABASE or DROP DATABASE failed and was rolled back part way through.
I've never heard of it happening in normal usage, though. If you
have a reproducible way of creating this situation I'd like to know
what it is.
>> The only work around so far
>> is to completely destroy the postgres instance and start over.
Vacuuming pg_database should fix it. The problem is that backend
startup is seizing on a dead pg_database tuple to tell it the OID
of the target database; vacuum out the dead tuples and you'll be OK.
regards, tom lane
PS: Memo to folks reporting problems: PLEASE quote any error message
you may be getting EXACTLY. It's a lot easier to look at the related
code if we can just grep for the message...
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Philip Warner | 2001-01-04 03:01:03 | Re: Please review TODO list |
| Previous Message | Barnes, Sandy (Sandra) | 2001-01-03 20:23:57 | V7.0 Error: OID ##### no longer exists in pg_databas e |