Joe Markwardt <joe@avsupport.com> writes:
>> The next thing to check is whether the OID mentioned in pg_trigger
>> (which I forgot already) is the same as the OID shown in pg_class, ie
>> select oid from pg_class where relname = 'pl02_status_table'.
> oid
> -----------
> 438094756
> (1 row)
> The oid that shows up in the error is 438095645.
> the oid for the pl02_status_table does not show up in any of the rows in
> pg_trigger that have the oid from the error in them.
Weird. You sure you didn't drop and recreate pl02_status_table? Table
oids don't change by themselves.
> The oid for the
> status table does show up in a few other rows in pg_trigger (attached
> file), which makes sense since its referenced in at least 2 other tables
> with foreign keys, and it still is being replicated by slony so it has
> that trigger on it as well.
Slony eh? You didn't mention that before; it does a few weird things IIRC.
You might try asking on the slony list whether anyone has any ideas.
regards, tom lane