On Mon, Sep 17, 2007 at 10:17:00PM -0700, Jason L. Buberel wrote:
> Thank you very much for the detailed response. In my case, the table is 
> not even part of a replication set, but I do need to bulk-load data into 
> the table in question on a regular basis...or at least I think I do. 

Well, one way to do this, I think, is to use a view of the underlying
tables.  You have more than one table involved at a time.  When you
perform your bulk load, you pull it into an empty table.  When you're
done, you change the view on all nodes, and presto! your new data
appears everywhere.

This is most useful in cases where you can also purge off old data,
so that you have some set of "rotor" tables, only one of which you're
loading at any one time.

> I have resisted implementing this at the application level, hoping to 
> come across a tool that would allow me to do this bi-directional 
> synchronization. Does any such beast exist, or is this ulimately a 
> do-it-yourself job?

It's do it yourself, but this sort of multimaster approach (with no
conflict resolution) is something people have proposed extending
Slony to solve.  You'd do something similar as the view approach
above.  The underlying tables would still need to be replicated,
however, which is different from the way you're doing it now.

A

-- 
Andrew Sullivan  | ajs at crankycanuck.ca
In the future this spectacle of the middle classes shocking the avant-
garde will probably become the textbook definition of Postmodernism. 
                --Brad Holland