Re: pg_dump: schema with OID xxxxxxxx does not exist - Mailing list pgsql-admin

From Tom Lane
Subject Re: pg_dump: schema with OID xxxxxxxx does not exist
Date
Msg-id 6801.1161967193@sss.pgh.pa.us
Whole thread Raw
In response to pg_dump: schema with OID xxxxxxxx does not exist  (Peter Neish <peter@austehc.unimelb.edu.au>)
Responses Re: pg_dump: schema with OID xxxxxxxx does not exist
List pgsql-admin
Peter Neish <peter@austehc.unimelb.edu.au> writes:
>  entity_browse_bid_seq |     37647910 |     1005 |      4 | t        |
> c       | t            | ,        | 37894096 |       0 | record_in |
> record_out | record_recv | record_send | i        | p          |
> f          |           0 |        -1 |        0 |               |
>  entity_browse         |     37647910 |     1005 |      4 | t        |
> c       | t            | ,        | 37894098 |       0 | record_in |
> record_out | record_recv | record_send | i        | p          |
> f          |           0 |        -1 |        0 |               |
> (2 rows)

> Does this mean that I can just delete these rows from this table for the
> dump to work? Will deleting these rows affect anything else?

These are composite-type entries, which means there should be
associated entries in pg_class --- with OIDs 37894096 and 37894098
respectively.  Did you check to see if those are still there?

We've seen scattered reports of table rowtype entries not being removed
when the owning table is dropped, but no one's ever been able to make a
reproducible case :-(.  Do you know anything about the history or usage
of these tables?

            regards, tom lane

pgsql-admin by date:

Previous
From: "Simon Kinsella"
Date:
Subject: Re: Postgres timestamps are out by one hour
Next
From: Peter Neish
Date:
Subject: Re: pg_dump: schema with OID xxxxxxxx does not exist