Re: PG9.1.2 -- 3 day old orphaned/non-responsive query -- pg_(cancel)(terminate)_backend has no effect - Mailing list pgsql-general

From Josh Kupershmidt
Subject Re: PG9.1.2 -- 3 day old orphaned/non-responsive query -- pg_(cancel)(terminate)_backend has no effect
Date
Msg-id CAK3UJREtFA8t7mCQwujoVu0vrB-Cd=dxsq-rSqYdDq=j=CoxAA@mail.gmail.com
Whole thread Raw
In response to PG9.1.2 -- 3 day old orphaned/non-responsive query -- pg_(cancel)(terminate)_backend has no effect  (Reid Thompson <Reid.Thompson@ateb.com>)
List pgsql-general
On Fri, Jan 13, 2012 at 12:02 PM, Reid Thompson <Reid.Thompson@ateb.com> wrote:

> The select query has been apparently orphaned by it's originating
> terminal or script or application.  pg_/cancel/terminate/_backend both
> return t, but the query does not cancel/terminate.   sigterm via the OS
> has no effect either.

You're sure you're killing the right backend (looks like the process
ID should be 32697, from what you posted)? What does strace tell you
the backend is doing?

> I'm not sure whether the vacuum is active, and just slow, or if it's
> spinning, or...?  I've taken no action toward the vacuum process.

Above you talk about the "select query has been apparently orphaned".
Are you trying to kill off the vacuum or the SELECT query?

Josh

pgsql-general by date:

Previous
From: Tomas Vondra
Date:
Subject: Re: indexes no longer used after shutdown during reindexing
Next
From: plasmasoftware net
Date:
Subject: how to make select for multiunit