the number one aim at the moment is to have "always-up-to-date" copy of our main DB with minial performance impact on replication (as I guess, single master - slave setup will work the best in this case). Eventually I it's likely that we'll want to unload the database server by splitting requests between two+ servers, and in this case having multi-master setup will be more convenient to have, I think. so at the point any of slony and pgcluster works for me, but before I start messing with any, I wanted to hear real users opninon about those (or different) packages :) > i think you need to be more specific about your replication requirements. > > async multi master is problematic in any case. it can be useful in certain > circumstances, but for generically duplicating a large database, it's generally > the wrong answer. > > you should probably focus on single master/multi slave setups, there are > useful solutions in that space. > > richard > -- > Richard Welty rwelty@xxxxxxxxxxxxxxx > Averill Park Networking > Java, PHP, PostgreSQL, Unix, Linux, IP Network Engineering, Security > "F=ma : it's not just a good idea, it's the law" > -- Vlad ---------------------------(end of broadcast)--------------------------- TIP 1: subscribe and unsubscribe commands go to majordomo@xxxxxxxxxxxxxx
