Fri Jan 12 07:04:17 PST 2007
- Previous message: [Slony1-general] Fwd: SLONY-I disadvantages
- Next message: [Slony1-general] update functions on 1.2.6 fails
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Hello, We are currently running a Postgres 8.1.3 database on an Ubuntu 5.10 server and recently implemented slony to replicate it. We're encountering an issue when including the sequences in our db schema in the replication set. When replicating only our tables (about 100 tables in total), performance and replication is smooth. However, when including the sequences in the replication process (about 14 sequences), this system load on our database server skyrockets after a couple of hours and the DB becomes unresponsive to any communication from our application. Without the sequences included in the set, the load average hovers around the 0.2 up to the 2.0 range when being stress tested, but when including them, the load rises to 10 and more after a few hours of normal operation, at which point the server eventually becomes unresponsive. While I realize it is warned that replicating sequences incurs a performance hit, I can't imagine such an extreme effect to be expected. Are there any configuration options you can suggest to us to explore and experiment with to deal with this before reporting anything on the Slony bug tracker? Thanks, Costa Galanis
- Previous message: [Slony1-general] Fwd: SLONY-I disadvantages
- Next message: [Slony1-general] update functions on 1.2.6 fails
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Slony1-general mailing list