Tue Sep 15 20:33:20 PDT 2009
- Previous message: [Slony1-general] Mixing slony versions in a cluster?
- Next message: [Slony1-general] uuid in postgresql on CentOS5
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
We are doing something very similar and it works fine. Two 8.4 nodes and one 8.3, with Slony 2.0.2. The slony build has to be built against the version of Postgres on the machine it will run on. So the easiest way is to have a single Slony 2.0.3 build against 8.4, and run all your slon processes on one of the 8.4 servers. That's how we run. Or if you prefer to run distributed slons, you can build a separate 2.0.3 Slony against 8.3 for the 8.3 server. You may get a warning about versions at startup, but it can be ignored so long as you don't start using 8.4-specific data features (not sure what they are). Hope that helps -gordon Dan Sugalski wrote: > > I've currently got a three node cluster in production running Slony > 2.0.2 and PG 8.3.7. It's ticking along pretty happily, and I'm > looking to toss a PG 8.4.1 node in to test the version upgrade. > > What I'm wondering is whether it's okay to install Slony 2.0.3 RC2 on > the new 8.4.1 node > -- View this message in context: http://www.nabble.com/Mixing-slony-versions-in-a-cluster--tp25464916p25465419.html Sent from the Slony-I -- General mailing list archive at Nabble.com.
- Previous message: [Slony1-general] Mixing slony versions in a cluster?
- Next message: [Slony1-general] uuid in postgresql on CentOS5
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Slony1-general mailing list