In both checkpointer.c and bgwriter.c, we do this before entering the
main loop:
/*
* Use the recovery target timeline ID during recovery
*/
if (RecoveryInProgress())
ThisTimeLineID = GetRecoveryTargetTLI();
That seems reasonable. However, since it's only done once, when the
process starts up, ThisTimeLineID is never updated in those processes,
even though the startup process changes recovery target timeline.
That actually seems harmless to me, and I also haven't heard of any
complaints of misbehavior in 9.1 or 9.2 caused by that. I'm not sure why
we bother to set ThisTimeLineID in those processes in the first place. I
think we did it when streaming replication was introduced because it was
an easy thing to do, because back then recovery target timeline never
changed after recovery was started. But now that it can change, I don't
think that makes sense anymore.
So, I propose that we simply remove those, and leave ThisTimeLineID at
zero in checkpointer and bgwriter processes, while we're still in
recovery. ThisTimeLineID is updated anyway before performing the first
checkpoint after finishing recovery, and AFAICS that's the first time
the value is needed.
- Heikki
--
Sent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-hackers
