Re: 8.3.7, `cache lookup failed` for a table

-----BEGIN PGP SIGNED MESSAGE-----
Hash: RIPEMD160

>> You think it is because slony is poking around pg_catalog.
>> schema, and it shouldn't , basically ?

> No, Slony 1.2.x pokes around in pg_catalog because in versions
> of postgres prior to 8.3 (which 1.2.x has to support) there was
> no built in way to disable the triggers.

This is not technically true. We had the same problem when
designing Bucardo. The canonical (and completely safe) way
is to do ALTER TABLE DISABLE TRIGGER. However, the locking
involved in such a call is too expensive for almost all
uses of a replication system, so both Slony and Bucardo made
the call to muck with the system catalogs instead. This does
lead to the occassional "OID not found" message, especially
on very busy systems.

It should also be noted that Bucardo and Slony disable rules
as well as triggers - something that cannot be done at the
SQL level. So the catalog tweaks are really necessary after
all in order to disable triggers *and* rules.

As pointed out, this is now completely solved with the
session_replication_role setting, fully supported by Bucardo,
and by Slony in the 2.x branch.

(The above glosses over a few details between the systems, but
it's mostly a moot point now as pre-8.3 systems are becoming
relatively rare)

- --
Greg Sabino Mullane greg(at)turnstep(dot)com
End Point Corporation http://www.endpoint.com/
PGP Key: 0x14964AC8 201005121002
http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8
-----BEGIN PGP SIGNATURE-----

iEUEAREDAAYFAkvqtbQACgkQvJuQZxSWSsjnqQCcDOTVMpCFowgLCR8EVG+V0LsD
pykAmNr6YTbtBRxHXB55ZXErevG07Js=
=LtbH
-----END PGP SIGNATURE-----