From: Tom Lane <tgl@sss.pgh.pa.us>
To: Ron <ronljohnsonjr@gmail.com>
cc: pgsql-admin@lists.postgresql.org
Subject: Re: Rollback after "Connection timed out"?
In-reply-to: <f90c644c-790a-46a9-ae95-1a10f6423d37@gmail.com>
References: <f90c644c-790a-46a9-ae95-1a10f6423d37@gmail.com>
Comments: In-reply-to Ron <ronljohnsonjr@gmail.com>
	message dated "Fri, 27 Oct 2023 09:40:43 -0500"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <1552282.1698417902.1@sss.pgh.pa.us>
Content-Transfer-Encoding: 8bit
Date: Fri, 27 Oct 2023 10:45:02 -0400
Message-ID: <1552283.1698417902@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1552283.1698417902%40sss.pgh.pa.us>
Precedence: bulk

Ron <ronljohnsonjr@gmail.com> writes:
> On Server_A, I ran:
> psql --host=blarge mydb -c "ALTER TABLE blah ..."

> After about 2 hours, it displayed this message:
> Could not receive data from server: Connection timed out
> Connection to server was lost

> And yet the ALTER TABLE (it's heavily inheritance-partitioned) is still 
> running on blarge.  Is it still working, or is it rolling back (which I'd 
> expect to be fast)?

It will probably run to completion if you leave it alone.  The
server would only notice loss of connection when it next tries to
send something to the client, which (barring error/notice
messages) would be the command completion report.

For next time, you might want to think about reducing the
TCP keepalive interval settings.

			regards, tom lane