Load balance Three Postgresql database (Primary, Secondery1 and Secondery2) Servers

Hi Experts,

I have somewhat large Postgresql database to manage and it's about 10GB after I made a dump and I believe it will take about 100GB space after it's populated on a hard disk. Currently I have implemented this on a Standby hot replication cluster. Two basic modules are running in this application and some reporting modules are directly connecting to the Secondary node(read only) and still there some SELECT queries coming in to Primary node(read/write) too.

Pgbouncer has implemented on both of these two database servers and currently operates on Transaction mode.  

I get too high I/O rate and Processing rates ( %I/O wait is around 15 ~ 20, Load leverage is around 8 ~ 10) in the peak times. (these two databases are running on SSD with 8 cores of 2.5 Ghz, two Blades, Ubuntu 12.04 server 64bit )

So I'm planning to add third read only server and implement a mechanism to balance the load across these three DB servers. Example, filter all the read requests on secondery1, secondery2  and write requests on the Primary.

Did any one tried this with pgpool-II, or what is your valuable recommendation on such a situation ?

Thanks a lot for your time !
Shakthi777Asked:
Who is Participating?
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

lcohanDatabase AnalystCommented:
"Two basic modules are running in this application and some reporting modules are directly connecting to the Secondary node(read only) and still there some SELECT queries coming in to Primary node(read/write) too. "

In my opinion it would be better to direct the traffic from the application to the desired instance and not from the backend database servers. In other words the app should have 2/3 connection strings (and app pools) to direct the queries against the desired "node" for the desired purpose.
0
Daniel WilsonCommented:
Another option is pl/Proxy.
https://wiki.postgresql.org/wiki/PL/Proxy

That is the solution Hannu Krosing developed to scale Postgres for Skype.

I haven't yet implemented a cluster in PostgreSQL, but PL/Proxy is the tool I expect to use ... later this year.
0

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
Shakthi777Author Commented:
I really wanted to thank for all your prompt responses,

#lcohan
This applications has implemented with a single connections string, any good approach on separating those ?

#Daniel Wilson
Partitioning is another option it's seems ?
0
Daniel WilsonCommented:
PL/Proxy involves a partitioning / sharding scheme.  The application knows nothing of the sharding.  It simply talks to the proxy.  The proxy knows which shard to contact .
0
Shakthi777Author Commented:
Thanks for the valuable comments !
0
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
PostgreSQL

From novice to tech pro — start learning today.

Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.