On Wed, Jan 5, 2011 at 2:09 AM, kakarukeys <kakarukeys@gmail.com> wrote:
> As requested, here are some output of the investigative queries, run
> when the problem occurred. I could see some locks there, but I don't
> know why the alter table add constraint takes so long of time.
It's pretty clear from the output you posted that it's waiting for a
lock, but you didn't include the full contents of pg_stat_activity and
pg_locks, so we can't see who has the lock it's waiting for. Tom's
guess upthread is a good bet, though.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company