Thread: Using Java for Replication :-)
This is not a rehash of the thread that Josh wanted killed, but more a pointer to a press release for "Daffodil Replicator", a java based, multi-database replication tool. From the press release (http://www.emediawire.com/releases/2004/11/emw179696.htm) "The released code includes all capabilities of the latest version (v1.4). These include bi-directional data synchronization, publish and subscribe model with server and client architecture, conflict detector and resolution algorithms and data synchronization between heterogeneous databases that support JDBC drivers, triggers, procedures and auto-increment columns." Sounds spiffy; I wonder how it stacks up against our family of postgresql specific solutions. if anyone has tried it out (or has time to try it out) please let us know. Robert Treat -- Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL
Robert, > This is not a rehash of the thread that Josh wanted killed, but more a > pointer to a press release for "Daffodil Replicator", a java based, > multi-database replication tool. From the press release > (http://www.emediawire.com/releases/2004/11/emw179696.htm) Cool, glad to have another option. There's another one, C-JDBC from Objectweb. It's a query-dispatch replicator, much like pgpool. Unfortunately, PG needs 2PC+XA support for the solution to work fully. -- Josh Berkus Aglio Database Solutions San Francisco
Josh Berkus wrote: > There's another one, C-JDBC from Objectweb. It's a query-dispatch > replicator, much like pgpool. Unfortunately, PG needs 2PC+XA > support for the solution to work fully. I talked to the ObjectWeb people last year and they seemed to say that it doesn't need 2PC. Where did you get your information? -- Peter Eisentraut http://developer.postgresql.org/~petere/
Peter, > I talked to the ObjectWeb people last year and they seemed to say that > it doesn't need 2PC. Where did you get your information? I'm on the C-JDBC mailing list, and have been chatting with ObjectWeb about getting 2PC implemented. The XA support is necessary to protect C-JDBC + Clustering transactions from failing on one server and succeeding on others; in other words, to guarentee a consistent state. You can use it now without that, but it makes things less dependable. -- Josh Berkus Aglio Database Solutions San Francisco