Re: Hot standby and removing VACUUM FULL

Greg Smith wrote:
> Heikki Linnakangas wrote:
>> So I guess what I'm asking is: Does anyone see any show-stoppers in
>> removing VACUUM FULL
> Here's the disclaimers attached to the new VACUUM REPLACE implementation
> from Itagaki:
>
> "We still need traditional VACUUM FULL behavior for system catalog
> because we cannot change relfilenode for them. Also, VACUUM FULL REPLACE
> is not always better than traditional VACUUM FULL; the new version
> requires additional disk space and might be slower if we have a few dead
> tuples."
>
> That first part seems like it would limit the ability to completely
> discard the current behavior.

For system catalog,s you could still use a utility like the one I
experimented with at
http://archives.postgresql.org/message-id/4AB15065.1000607@enterprisedb.com.
Essentially, do a bunch of dummy UPDATEs on the rows that you want to
move. It can cause serialization errors in concurrent updaters, like any
UPDATE, but I think it would be good enough for the narrow remaining use
case of shrinking system catalogs.

--
Heikki Linnakangas
EnterpriseDB http://www.enterprisedb.com