Horton

Re: NOTIFY does not work as expected

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

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andres Freund <andres@anarazel.de>
Cc: Andrey <parihaaraka@gmail.com>, Jeff Janes <jeff.janes@gmail.com>, Pg Bugs <pgsql-bugs@postgresql.org>
Date: 2018-10-20T00:27:20Z
Lists: pgsql-bugs 
Andres Freund <andres@anarazel.de> writes:
> I find the +1, 0, -1 pretty confusing and less clear than what was there
> before. If we do something like it, could we introduce an enum or macros
> for it?

Meh.  Shall I just go back to the blocked/not blocked approach?  There
was a need for the 3-way argument at one point while I was fooling with
it, but once I decided that it'd be a good idea to re-set the latch on
the way out, there wasn't anymore.

			regards, tom lane

Commits

  1. Client-side fixes for delayed NOTIFY receipt.

  2. Server-side fix for delayed NOTIFY and SIGTERM processing.

  3. Introduce and use infrastructure for interrupt processing during client reads.