Hi,

This project is really cool. We have multiple applications whose
datasources are now utilizing Phoenix. Here is my critical question:

Our applications and some 3rd party projects we depend on are already using
some form of connection pools. I read from one of the previous posts saying
that we DO NOT NEED to use connection pools but is there a danger
continuing with our existing codebase? I believe that we should depend on
interfaces but not implementations. So, I want to assume that an already
running application utilizing connection pools should easily be migrated by
updating the JDBC url to point to phoenix url. Otherwise, a lot of changes
need to be made on both our applications and also on 3rd party dependencies
like Saiku. As an example, Saiku by itself uses a kind of HashMap to store
connections instead of opening them again and again.

I am looking forward to getting some direction on this.

Thanks a lot!

Reply via email to