Hello all!
Still haven't found any solution to this problem.
Having a Postgres 9.0 with assynchronous streaming replication to a
hot-standby slave, both with CentOs 5.6, how can I use both DB instances
for query load balancing?
I've tried with pgPool-II but the pooling mechanism is disruptive with
the existing pool on the application servers.
Has anyone had this issue before? Any suggestions?
Best regards,
Paulo Correia
On 13/04/12 14:38, Paulo Correia wrote:
> Hi!
>
> I have a Postgres 9.0 on CentOs 5.6 with a functional asynchronous
> streaming replication to a hot-standby slave (over a TCP connection).
>
> At the moment, the slave only makes the replication, but it accepts
> read-only queries.
>
> I need to load-balance the DB requests to both servers and was trying
> pgPool-II but it seems it has some problems if we already have
> connection pooling elsewhere.
> I have some application servers that already have their own connection
> pool functionalities and I wonder if anyone found a solution for that.
>
> Imagine I have some different deploys on a Glassfish or a Tomcat
> server, using the server connection pooling facilities and from that
> we would access the database, or the balancer.
>
> Has anyone managed to do this with pgPool-II? Any other options?
>
> Best regards,
> Paulo Correia
>
>