On Mon, Jan 18, 2016 at 5:50 PM, Thomas Kellerer <spam_eater@gmx.net> wrote:
> With all the problems I have seen (in Oracle and Postgres) I think that
> maybe a better solution to this problem is to make the planner fast (and
> reliable) enough so that plan caching isn't necessary in the first place.
>
> However I have no idea how feasible that is.
The problem is that the floor is already littered with
potentially-very-beneficial query planning ideas that got discarded
because they would add too many cycles to planning time. Despite
that, planning time is a killer on some workloads. So right now we've
got workloads where we plan too much and workloads where we plan too
little. Argh.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company