Hi all, Some time ago I said I'd prepare a little script to help with bulk loads under Slony-I. Attached is such a script.
PLEASE READ THE DOCS before you use this. This thing is _dangerous_.
It's intended to solve a specific problem. It is _really really not_
intended to solve generic load-and-replace issues. Even more, there
are plenty of places where it can completely break replication in
your Slony cluster on a table. I strongly suggest you use this on a
table that is isolated in its own set. Even there, things might
break.
I mean it. It's safe only in case you're completely willing to throw
away all the table data and start over. It makes changes under the
radar. Please don't ask me for support when you use this on
production data. No warranty. And so on. It hasn't been tested; it
might well break in your situation. Please test it before you use
it.
All of that said, this can be used if you are adding a table to
replication, and need a _lot_ of data to show up in it, and can be
sure that you don't care if you have to start over. It avoids
the problem where you load everything on the origin, and then ship
all that data retail over your replication connection.
It does not disable or enable indexes; you're supposed to do all that
stuff yourself in a wrapper script, or pay the cost of updating the
index retail.
I hope others find this useful (and that tarball attachments make it
through on the mailing list!)
A
--
Andrew Sullivan | [EMAIL PROTECTED]
The plural of anecdote is not data.
--Roger Brinner
slony_bulkload.tar.bz2
Description: Binary data
_______________________________________________ Slony1-general mailing list [email protected] http://gborg.postgresql.org/mailman/listinfo/slony1-general
