Instead of DBA_JOBS_RUNNING, I've relied on 'THIS_DATE' being not null to
indicate that my job is running. This suggests that the job finished in a
short time. This is supported by some logging we perform at the start and
end of our procedure (not shown).
I will add DBA_JOBS_RUNNING to my observations, to see if this agrees with
THIS_DATE being null.
Job is running:
WHAT
LAST_DATE NEXT_DATE THIS_DATE DECODE(TH
-------------------- -------------------- -------------------- ---------
BEGIN manhatten_com.poll_label; END;
14-JUL-2006 12:50:31 14-JUL-2006 12:50:34 14-JUL-2006 12:50:36 EXECUTING
.... job finishes (I think) as THIS_DATE is now NULL. Next_date is set to
three seconds in the future.
LAST_DATE NEXT_DATE THIS_DATE DECODE(TH
-------------------- -------------------- -------------------- ---------
BEGIN manhatten_com.poll_label; END;
14-JUL-2006 12:50:36 14-JUL-2006 12:50:39
.... and finally, it appear to have executed again. (I didn't capture it
executing, as it is quick).
WHAT
LAST_DATE NEXT_DATE THIS_DATE DECODE(TH
-------------------- -------------------- -------------------- ---------
BEGIN manhatten_com.poll_label; END;
14-JUL-2006 12:54:01 14-JUL-2006 12:54:04
--
http://www.freelists.org/webpage/oracle-l
Received on Fri Jul 14 2006 - 10:12:39 CDT