On Fri, 19 May 2006, Tom Lane wrote:
Jeff Frost <jeff@xxxxxxxxxxxxxxxxxxxxxx> writes:
Do you think the postmaster on 5432 is trying to archive the other
postmaster's WAL files somehow?
Not as long as they aren't in the same data directory ;-). What Simon
was wondering about was whether an archiver process had somehow been
left over from a previous incarnation of the test postmaster. The thing
to do is look through "ps auxww" (or local equivalent) and see if you
see more than one thing calling itself an archiver process.
(Whether or not this explains Jeff's problem, it definitely seems like
a failure mode that we need to guard against. We go to great lengths
to prevent a new postmaster from starting when there are still live
backends from a previous postmaster, but I don't think that interlock
is effective for the archiver.)
Well now, will you look at this:
postgres 20228 1 0 May17 ? 00:00:00 postgres: archiver process
postgres 20573 1 0 May17 ? 00:00:00 postgres: archiver process
postgres 23817 23810 0 May17 pts/11 00:00:00 postgres: archiver process
23810 is the running postmaster:
postgres 23810 1 0 May17 pts/11 00:03:01 /usr/local/pgsql-8.1.3/bin/postm
do you think that got left around the last time I did a pg_ctl restart?
I guess I can stop my re-run of the test after stopping the /var/lib/pgsql
postmaster.
--
Jeff Frost, Owner <jeff@xxxxxxxxxxxxxxxxxxxxxx>
Frost Consulting, LLC http://www.frostconsultingllc.com/
Phone: 650-780-7908 FAX: 650-649-1954
