On Tue, Mar 7, 2023 at 10:19 PM Bryn Llewellyn <bryn@yugabyte.com> wrote:
Why not err on the side of caution and (I trust) guaranteed currency of each session's in-memory representation of a PL/pgSQL program with the environment in which it executes?
After all, you add a column in order to use it. And this means that at the very least client-side code must be changed to do this. And this means quiescing use of the application and then re-starting it with new behavior. Is re-starting the connection pool before opening up the new app for use so expensive that it's worth trying to reason when it might be safe to avoid this re-start?
True. I was a bit hasty in forming an opinion on an operational aspect like that. That just isn't something I typically think about.
