On 2013-06-07 19:09, Fred&Dani&Pandora&Aquiles wrote:
> I asked a while ago in this group about the possibility to implement a
> parallel planner in a multithread way, and the replies were that the
> proposed approach couldn't be implemented, because the postgres is not
> thread-safe. With the new feature Background Worker Processes, such
> implementation would be possible? If yes, do you can see possible
> problems in implement this approach, for example, the bgprocess can't
> access some planning core functions like make_join_rel, acess them in
> parallel and so on. I want start a research to work in a parallel
> planner in postgres, I succeeded in in the DBMS H2, but my first
> option still is the postgres, and any help is welcome.
The topic has been researched and experimented with on PostgreSQL 8.3,
described in a vldb-2008 paper called Parallelizing Query Optimization,
available on http://www.vldb.org/pvldb/1/1453882.pdf
regards,
Yeb
PS: apologies for redundant posting.