[Pgpool-general] Pgpool and/or Slony

[Brad Nicholson]

On Mon, 2007-01-15 at 14:28 -0500, Andrew Sullivan wrote:
> On Fri, Jan 12, 2007 at 02:31:29PM -0500, Brad Nicholson wrote:
> > With slony, now() will get fired on the master, update foo with the
> > timestamp, and then replicate that exact timestamp to the subscribers.
> > 
> > With pgpool you will get the value of now() on each server, which are
> > very likely not going to match.
> 
> I believe this particular limitation is no longer true (if ever it
> was), in that pgpool has some intelligence about translating this
> sort of data in some cases.  The key thing, however, is that there
> _are_ cases where statement-based replication has some difficulties,
> and some of them can't be worked around in principle.

Are you talking pgpool-I or pgpool-II?

pgpool docs clearly state:

"Please note that certain queries are physically dependent to a server
or non-deterministic. These include random functions, OID, XID, and
timestamps may not be replicated in exactly same value among two
servers. This is because query results for these are local to each
PostgreSQL server."

-- 
Brad Nicholson  416-673-4106
Database Administrator, Afilias Canada Corp.