If your interval is the string you provided, I would expect the next run date to be 9/5/2006 00:30.
Selecting the string from dual provides that date. If the job failed, then the scheduler will set the next run date to try again, ignoring the interval. I forget how the algorithm works, it is in the docs.
As an aside, you might consider just using date math instead of the to_date string handling stuff.
eg. select trunc(add_months(sysdate,1),'mm') + 4 + (.5/24) from dual;
Jared
Oracle 8.1.7, Solaris 2.9. We have a job that is not running properly. Could that be why it is unexpectedly trying to execute at a different time when the next scheduled execution time should have been August 5 2006. Also the next run date is now showing as August 2 at 4 AM. We are trying to understand why it should have done so even though everything else is the same in the job properties. The properties show the interval as TO_DATE('5-'|| TO_CHAR(ADD_MONTHS(sysdate,1),'MON-YYYY'), 'DD-MON-YYYY') + 00.50/24
We recently executed the following to change the package - the package fails at the same time each day - the time of deployment - not the scheduled time. Also, around this same time the status of the job was switched from broken (for deployment) to fixed:
BEGIN DBMS_JOB.WHAT(342, 'cpsdataminer_pkg_p3.cpseompenaltycalc;'); END; /
------------------------------ Do you Yahoo!? Next-gen email? Have it all with the all-new Yahoo! Mail Beta.<http://us.rd.yahoo.com/evt=42241/*http://advision.webevents.yahoo.com/handraisers>
-- Jared Still Certifiable Oracle DBA and Part Time Perl Evangelist