Horton

Re: [sqlsmith] ERROR: badly formatted node string "RESTRICTINFO...

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andreas Seltenreich <seltenreich@gmx.de>
Cc: pgsql-hackers@postgresql.org
Date: 2017-04-11T16:11:34Z
Lists: pgsql-hackers 
Andreas Seltenreich <seltenreich@gmx.de> writes:
> I see the above ERROR logged a lot when testing master at eef8c0069e
> with a postgres_fdw around.  Below is a recipe to reproduce it on top of
> the regression DB.

I've pushed a fix that should get you past that problem, although
I suspect we still have some issues with treatment of parallel-unsafe
subplans.

			regards, tom lane

Commits

  1. Avoid transferring parallel-unsafe subplans to parallel workers.

  2. Mark finished Plan nodes with parallel_safe flags.

  3. Simplify handling of remote-qual pass-forward in postgres_fdw.

  4. Handle restriction clause lists more uniformly in postgres_fdw.

  5. Allow parallel workers to execute subplans.