Rhhh Lin <ruanline...@hotmail.com> writes: > So I decided to intervene and terminate some active sessions (Which were > active and not waiting) which were all querying this table with a similar > query, by using pg_backend_terminate(pid). Some of these sessions were in an > active state for a few days executing this query. However, the termination of > the PIDs [The command succeeded and returned 't']has not closed down the > sessions. They are still present, and still active and still executing. Those > current sessions are for whatever reason completely "hung", locking out > exclusive access to the table (from vacuum) and cannot be killed. I have not > yet resorted to trying to kill the process PID's from the OS perspective. > Instead, I will try to schedule an outage over the next few days, restart the > server, implement a vacuum execution and see where I am at then.
Well, that's pretty interesting in itself. Any chance of attaching to one of those unkillable backends with gdb and getting a stack trace? https://wiki.postgresql.org/wiki/Generating_a_stack_trace_of_a_PostgreSQL_backend regards, tom lane -- Sent via pgsql-general mailing list (pgsql-general@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
