Tom Lane [tgl@xxxxxxxxxxxxx] wrote:
Well, that's certainly a sufficient reason, if perhaps not the only
reason. Dropping ts_defects_20090227 will require removal of FK triggers
on ts_transets, and we can't do that concurrently with transactions that
might be trying to fire those triggers.
Now admittedly, it would probably be sufficient to take ExclusiveLock
rather than AccessExclusiveLock when removing triggers, since we do not
have triggers ON SELECT. Right now though, we just take
AccessExclusiveLock for most any DDL on a table. There was a patch
submitted last fall to reduce DDL locking in some cases, but it hasn't
been reworked to fix the problems that were pointed out (and I
disremember if it addressed DROP TRIGGER in particular anyway).
regards, tom lane
Thanks for furthering my understanding of postgres (and probably other
SQL servers as well). I can fix this problem easily.
Brian
--
Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-performance
