Horton

Re: pg_recvlogical prints bogus error when interrupted

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Cc: Tristan Partin <tristan@neon.tech>, Kyotaro Horiguchi <horikyota.ntt@gmail.com>, andres@anarazel.de, pgsql-hackers@postgresql.org
Date: 2023-07-20T00:34:26Z
Lists: pgsql-hackers 
On Wed, Jul 19, 2023 at 01:33:15PM +0530, Bharath Rupireddy wrote:
> I think the delay is expected for the reason specified below and is
> not because of any of the changes in v5. As far as CTRL+C is
> concerned, it is a clean exit and hence we can't escape the while(1)
> loop.

Yes, that's also what I am expecting.  That's costly when replaying a
large change chunk, but we also want a clean exit on a signal as the
code comments document, so..
--
Michael

Commits

  1. Fix pg_recvlogical upon signal termination