https://pfspipe.ipmu.jp/jira This file is an XML representation of an issue en-us 8.3.4 803005 13-09-2019 https://pfspipe.ipmu.jp/jira/browse/INSTRM-121 Instrument control development <p>The archiver lost its connection to the database, due to a (security?) update made to the database server. I don't think it ever tried to reconnect, but kept caching new rows after the failure in the attached log segment.</p> <p>I don't think there is a trivial way to restart, but the hint at <a href="https://github.com/psycopg/psycopg2/issues/419" class="external-link" rel="nofollow">https://github.com/psycopg/psycopg2/issues/419</a> is probably good enough to implement from.</p> INSTRM-121

Archiver needs to handle database restarts

Task Normal Open Unresolved cloomis cloomis Fri, 9 Jun 2017 19:28:16 +0000 Tue, 9 Feb 2021 15:35:35 +0000 ics_archiver 0 1 <p>The archiver keeps files of unapplied updates in a /tmp/archiver-$PID/directory. With postgreql, those are suitable for <tt>COPY $table FROM '$file' CSV QUOTE ''''</tt> commands, which are exactly what the archiver uses.</p> <p>Since the archiver was complaining about the lost connection, I assumed I would have to do that semi-manually. But when I killed the archiver it applied all the updates itself.</p> <p>So I think we just to make the reconnect logic a bit smarter, and/or add an interrupt handler to force reconnections and updates.</p> <p>Bump. Recent database work caused reconnection and other issues.</p> <p>The archiver does respond well to SIGINTs: it unwedges itself and and flushes all pending transactions from the transaction log files. So it already knows how to to the right things.</p> <p>It just needs to do that automatically.</p> Development Rank 0|08c000: