[Pgpool-general] pgpool, slony and hibernate?
Martin Fandel
martin.fandel at alphyra.de
Thu Nov 2 09:11:40 UTC 2006
Hi,
I've don't tested hybernate, so I only can view a short list of the
main pros/contras (for my situation, everyone has a different
environemnt an a different situation for evaluating) about slony
and pgpool.
Slony
pros:
- every table has a serial / pkey for comparing the nodes
- possible to replicating parts of dbs to different nodes
contras:
- it's difficult in environments with many ddl changes
- no merge replication
- no load balancing mode!
- no automatic failover function (must use drop node / add
node / failover)
- difficult with more than 2 nodes (its working fine but if
anybody not knows what to do in a fail state, its dangerous)
pgpool
pros:
- easy to install
- load balancing and replication
- possible to replicate parts of tables to different nodes
- failover function
- really new in development i think this is a pro :)
contras:
- no comparing between nodes
- no merge replication
Have you tried PGCluster? It's a build in replication system. I've
tested it shortly, but I think for now pgpool-II is the best solution.
All projects are very very ingenious!
Regards,
Martin
Am Mittwoch, den 01.11.2006, 12:13 -0800 schrieb Jeff Frost:
> Is anyone out there using pgpool, slony and hibernate in a load balancing
> scenario? If so, can you relate any gotchas?
>
More information about the Pgpool-general
mailing list