On Fri, 2009-08-14 at 11:36 -0700, John R Pierce wrote:
> Martin Spinassi wrote:
> > So, what you suggest is:
> >
> > |---------> Master <---|
> > pg_pool2 ---| |----- Slony
> > |---------> Slave <---|
> >
> >
> > Can't figure out if this is gonna work, configuring pg_pool2 to read
> > both, making slony replicate, and app directly to Master (or pg_pool2
> > can be configured to write just in one of both?).
> >
>
> something like that.
>
> see the pgpool-II tutorial at
> http://pgpool.projects.postgresql.org/pgpool-II/doc/tutorial-en.html
>
> note it also can do replication, but it does it at the front end rather
> than the back end like slony. each has its advantages and disadvantages.
> (I've also heard, repeatedly, that pgpool-II's replication isn't that great)
>
Thanks John.
I'll try sqlrelay as mentioned by sramirez (sorry, no sign at the
email), and pg_pool2 as you suggest. Couldn't find any good
documentation of dblink, but think I can read the code if the first two
options doesn't work as expected.
Once again, thanks @all for you answers!
Cheers
Martin