3

I'm planning to migrate from MySQL to PostgreSQL, but worried about the apparent luck of replications solutions there.

Are there any current replication techniques that can be used out of box?

Heartbeat/DRBD would probably work, but I'm looking for approaches that are more integrated in PostgreSQL itself.

SyRenity
  • 3,159
  • 11
  • 55
  • 79
  • i like this one: 'the replication options for PSQL are agricultural to say the least' [ from http://serverfault.com/questions/11137/postgresql-replication ] ;-] – pQd Jan 05 '10 at 19:27
  • Yep, I seen this as well. Not very encouraging :( – SyRenity Jan 06 '10 at 11:56

1 Answers1

4

For available replication solutions, see for instance http://wiki.postgresql.org/wiki/Replication,_Clustering,_and_Connection_Pooling.

But no, there is nothing at the moment that works "out of the box".

PostgreSQL 8.5, currently in development, will contain a built-in replication feature that is known under the name "Hot Standby". See for example http://www.postgresql.org/about/news.1172.

Peter Eisentraut
  • 3,575
  • 1
  • 23
  • 21
  • This looks interesting - will the switch be automatic, and the "hot standby" node will become the master? Also any ETA for the 8.5 stable? Thanks. – SyRenity Jan 06 '10 at 11:57
  • No, the hot standby feature only arranges the replication of the data and receives notice when the slave should become the master. But you will have to trigger the switching yourself, using Heartbeat or UCARP or something like that. It's not exactly plug-and-play. – Peter Eisentraut Jan 06 '10 at 21:37