Re: Disabling DBMS_SCHEDULER

From: Rich Jesse <rjoralist_at_society.servebeer.com>
Date: Tue, 4 Mar 2008 15:35:14 -0600 (CST)
Message-ID: <51973.12.17.117.251.1204666514.squirrel@12.17.117.251>


> No. In my experience, setting job_queue_processes to zero only stops
> the dbms_jobs jobs from running.
>
> Daniel

Once upon a time, I had an SR open about this. There is this procedure:

dbms_scheduler.set_scheduler_attribute('SCHEDULER_DISABLED', 'TRUE');

...but I was not able to get it to work reliably in 10.1.0.5.0 on AIX. Even with it disabled, jobs inexplicably ran. What I ended up doing was to create a stub function that checked a status. This status is manually set by me (e.g. value in a single-row table or perhaps something "automatic" like instance name). Per protocol, every procedure in a DBMS_SCHEDULER job must first call this function. That way, I can emulate the disabled scheduler attribute as it was intended to be used.

It's a bit of a hack, but better for me than attempting to get a patch to fix the scheduler bug. Of course, YMMV, as my problem stemmed from testing RMAN recovery of our production database on a test node. And since the production jobs ran in this test recovery database, they also sent email, confusing the recipients. My function, however, is simple enough to work across multiple versions of Oracle.

HTH! GL! Rich

--
http://www.freelists.org/webpage/oracle-l
Received on Tue Mar 04 2008 - 15:35:14 CST

Original text of this message