Re: dbms_scheduler jobs stop running after reboot.

Hi Andy,

The job_queue_process is set to 50. Only the aq$ jobs (50+) are running in dba_scheduler_running_jobs. None of the in_house developed jobs ( 5 jobs fire every 5 min run on on 100 + schema) were running.

one cjq0 process. The jobs are not broken or disabled, Just stopped to run.

The Oracle is 12.1.2.0 on Linux.

Thanks,

Joan

On Thu, Feb 28, 2019 at 1:53 PM Andy Sayer <andysayer_at_gmail.com> wrote:

> What’s your job queue processes set as?
> Is there anything in dba_scheduler_running_jobs? We’ve had issues where
> dba_scheduler_running_jobs reports some jobs as running but with no
> session, we stopped them and the problem went away.
>
> Is there anything special about these jobs? Are they chains? With
> subchains?
>
> You said CJQN was running, what is it doing? When did it last do something?
>
> Thanks,
> Andy
>
> On Thu, 28 Feb 2019 at 18:30, Martin Berger <martin.a.berger_at_gmail.com>
> wrote:
>
>> Hi Joan,
>> It would be nice if you can tell us the version or your DB.
>> I am aware of events 27401 and 27402 regarding the scheduler. I don't
>> know a lot about 27401, but had some practical experience with
>> alter system set events '27402 trace name context forever, level 65535';
>> .
>> If you can set this and see after some minutes what's the content of cjq0
>> tracefile, maybe this help you to dig into the problems core.
>>
>> hth
>> Martin
>>
>>
>> --
>> Martin Berger Oracle ♠
>> martin.a.berger_at_gmail.com _at_martinberx <https://twitter.com/martinberx>
>> ^∆x http://berxblog.blogspot.com
>>
>>
>>
>> Am Do., 28. Feb. 2019 um 17:57 Uhr schrieb Joan Hsieh <
>> joanhsieh08_at_gmail.com>:
>>
>>> Hi List,
>>>
>>> We planned to increase the SGA_target from 50G to 64G. Ended up we
>>> had to change back to 50G due to the page size was not supported.
>>> However, after the database rebooted all of the scheduler jobs
>>> stopped to execute. (There are more than 500 jobs, fired every 5 min
>>> in application)
>>>
>>> The last_start_date in dba_scheduler_jobs was the time prior to the
>>> database rebooting, which was 2/23 9pm, and the state was SCHEDULED.
>>>
>>> The job queue is 50, cjg0 is running. we don't use oracle OIM.
>>>
>>> The developer had tried to disable/enable the scheduler. They had
>>> also tried to use exec DBMS_SCHEDULER.run_job to test. The job ran,
>>> but did not fire after 5 min for the next run.
>>>
>>> The database is 4 TB in size. Previously the database had bounced
>>> several times, but it had never caused any scheduler jobs issue
>>> before. We are trying to find any resource and had opened an Oracle
>>> SR for this issue. We just wanted to find out if you have have any
>>> insight.
>>>
>>> Much thanks in advance.
>>>
>>> Joan
>>>
>>

--
http://www.freelists.org/webpage/oracle-l