On Mon, 2006-05-22 at 15:33 -0400, Andrew Sullivan wrote: > On Mon, May 22, 2006 at 10:48:46AM -0500, Jim C. Nasby wrote: > > > It's tricky, but you can often do it without application changes, as > > > long as you never do something dumb like SELECT *. > > > > Maybe I missed something, but wasn't the OP trying to replicate data > > from slaves back up to the master? > > I didn't have that impression: I had the impression of many remote > machines, and one central one that shared their data. If this is to > be two-way, then it does indeed become more tricky (although the same > general strategy might be made to work).
I've had good luck with replicating single partitions (inherited tables) to various locations in odd combinations with inheritance taking care of the re-assembly. -- _______________________________________________ Slony1-general mailing list [email protected] http://gborg.postgresql.org/mailman/listinfo/slony1-general
