Re: src/test/subscription/t/005_encoding.pl is broken

Michael Paquier <michael.paquier@gmail.com>

From: Michael Paquier <michael.paquier@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Robert Haas <robertmhaas@gmail.com>, Andres Freund <andres@anarazel.de>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2017-09-20T05:32:56Z
Lists: pgsql-hackers

Attachments

On Wed, Sep 20, 2017 at 12:33 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> That would indicate that something isn't ever retrying the worker
> start; but if that's the case, how is it that we get through the
> other subscription tests with my random-failure patch in place?

I have been able to dig into this issue further, and the problem is
indeed in the wait logic of 005_encoding.pl. It is important to wait
for the initial sync of the subscriber to happen. There is no need to
incorporate the additional wait query in wait_for_caught_up() as well.
Please see the attached which fixes the stability problems for me even
after forcing failures in launcher.c.
-- 
Michael

Commits

  1. Fix instability in subscription regression test.