DBMS_SCHEDULER, schedule_limit
Oracle 10g RAC on AIX 5.x
we have a job run periodically with DBMS_SCHEDULER, with each occurence timed at 10 minutes. Normally the job runs take less than a minute, however during one of data feeds, one of the occurence took more time (around 25 minutes) & since it was more than the repeat_interval time (10minutes) & the schedule_limit attribute was set to null, my understanding is next occurrence should run after the job which took 25 minutes completed.
Observation---> The next occurence did take place after the 25 minute job, however the 2nd preeceding job was skipped & 3rd, 4th... & all jobs from then on were as per the schedule & ran successfully.
Question----> why was the 2nd subsequent job skipped?
I think the dba_scheduler_job_log view should have an entry of the skipped job, it doesn't, where can I get details about the skipped job, is there any other view? also is there a behavior of dbms_scheduler which has caused this job to be skipped?
Why did the job take 25 minutes? Is it because of the data feeds, or any suggestions on how this can be investigated?
Please let me know if you need more details on understanding this question, thanks & look forward to your suggestions.