Re: draft RFC: concept for partial, wal-based replication

[Craig Ringer <craig(at)postnewspapers(dot)com(dot)au>]

On 30/11/2009 11:07 PM, Tom Lane wrote:
> Craig Ringer<craig(at)postnewspapers(dot)com(dot)au>  writes:
> > Just a side note: in addition to its use for partial replication, this
> > might have potential for performance-prioritizing databases or tablespaces.
>
> > Being able to separate WAL logging so that different DBs, tablespaces,
> > etc went to different sets of WAL logs would allow a DBA to give some
> > databases or tablespaces dedicated WAL logging space on faster storage.
>
> I don't think this can possibly work without introducing data corruption
> issues.  What happens when a transaction touches tables in different
> tablespaces?  You can't apply the changes out-of-order.

Argh, good point, and one that should've been blindingly obvious.

At a database level something like that may still be handy, though I 
haven't the foggiest how one would handle the shared system catalogs.

--
Craig Ringer