Hello Team,
Could you guys please suggest the best solution for postgresql replication
setup with load balancing ?
Better have the ability to control where read operation goes by tables.
Let's "select table A" goes only to master, "select table B" goes only to
save, "select table C" goes to master and slave in turns
We have found that pgpool supports load balancing with many other feature
like connection pooling, replication, online recovery etc.
However, it doesn't support connection pooling with database multiplexing as
given below :
=============
test_db1 = host=master= port=5433= dbname=test_db1
test_db2 = host=slave= port=5433= dbname=test_db2
===========
So we tried to implement following setup:
1. Application Client Connects to PgBouncer
2. PgBouncer forwards request to a PgPool for that cluster
3. PgPool forwards Request to the PostreSQL server (master or slave)
4. PostgreSQL responds to the request
If the above architecture is possible and is the best, could you guys please
advice on how this can be implemented (backend-settings) ?
Thank you,
Abhilash
System Enginner
--
View this message in context:
http://postgresql.1045698.n5.nabble.com/Please-advice-a-best-solution-for-postgresql-replication-with-load-balancing-tp5792675.html
Sent from the PostgreSQL - admin mailing list archive at Nabble.com.