Re: pg_recvlogical prints bogus error when interrupted

Andres Freund <andres@anarazel.de>

From: Andres Freund <andres@anarazel.de>
To: Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>
Cc: Kyotaro Horiguchi <horikyota.ntt@gmail.com>, pgsql-hackers@postgresql.org
Date: 2022-10-27T23:11:39Z
Lists: pgsql-hackers
Hi,

On 2022-10-24 08:15:11 +0530, Bharath Rupireddy wrote:
> I came up with the attached v2 patch, please have a look.

Thanks for working on this!


> +		/* When we get SIGINT/SIGTERM, we exit */
> +		if (ready_to_exit)
> +		{
> +			/*
> +			 * Try informing the server about our exit, but don't wait around
> +			 * or retry on failure.
> +			 */
> +			(void) PQputCopyEnd(conn, NULL);
> +			(void) PQflush(conn);
> +			time_to_abort = ready_to_exit;

This doesn't strike me as great - because the ready_to_exit isn't checked in
the loop around StreamLogicalLog(), we'll reconnect if something else causes
StreamLogicalLog() to return.

Why do we need both time_to_abort and ready_to_exit? Perhaps worth noting that
time_to_abort is still an sig_atomic_t, but isn't modified in a signal
handler, which seems a bit unnecessarily confusing.

Greetings,

Andres Freund



Commits

  1. Fix pg_recvlogical upon signal termination