Re: [PERFORM] The state of PG replication in 2008/Q2?

2008-08-21 Thread Mathias Stjernström
I Agree with Robert but i never heard of Cybercluster before. Does anyone have any experience with Cybercluster? It sounds really interesting! Best regards, Mathias Stjernström http://www.pastbedti.me/ On 22 aug 2008, at 08.18, RW wrote: My company finally has the means to install a new

Re: [PERFORM] The state of PG replication in 2008/Q2?

2008-08-21 Thread Mathias Stjernström
with those slony scripts. My experience is that automatic handling of DDL changes is a very important feature of a replication system of curse not in all systems but in many. I am also very interested in the WAL replication that David Lang asked about. Best regards, Mathias Stjernström http

Re: [PERFORM] The state of PG replication in 2008/Q2?

2008-08-21 Thread Mathias Stjernström
23.04, Andrew Sullivan wrote: On Thu, Aug 21, 2008 at 10:53:05PM +0200, Mathias Stjernström wrote: For Master-Slave replication i think that Slony http://www.slony.info/ is most up to date. But it does not support DDL changes. This isn't quite true. It supports DDL; it just doesn't su

Re: [PERFORM] The state of PG replication in 2008/Q2?

2008-08-21 Thread Mathias Stjernström
Hi Dan! Its true, many of the replication options that exists for PostgreSQL have not seen any updates in a while. If you only looking for redundancy and not a performance gain you should look at PostgreSQL PITR (http://www.postgresql.org/docs/8.1/static/backup-online.html ) For Master-S

Re: [PERFORM] Using PK value as a String

2008-08-12 Thread Mathias Stjernström
Hi! We use normal sequences to generate id's across multiple nodes. We use the "increment" parameter for the sequence and we specify each node to increment its sequence with for example 10 and the the first node to start the sequence at 1 and the second at 2 and so on. In that way you ge