Postgres Replication
Posted 6:32 AM

There's a lot of talk about the lack of built-in replication for postgres. I have been using the "pg_comparator" method for slave updates.

This is essentially similar to an "rsync" for database tables. It works extremely well and has a low configuration overhead. You can use it to scale a single table out to dozens of servers, even choosing which fields to replicate.

It's also easy to install, configure and maintain - even code for.

Personally, I like it better than the builtin solution that they are discussing on postgres's lists. I wish a diff-patch shipping solution was in the running. For a lot of us, that's exactly what we need.

Labels: programming

[View/Post Comments] [Digg] [Del.icio.us] [Stumble]