cancel
Showing results for 
Search instead for 
Did you mean: 

"Scheduled" job status

Former Member
0 Kudos

Hi All,

There are sometimes some jobs that stay in the scheduled status for a while without any obvious reason: there is no Time Windows concerns and no RFC agent dependances. I have to run the "run job now" command to make it execute instantaneously.

Does anyone know the reason of this behavior ?

Thanks in advance for your help!

Regards,

Yi Jiang

Accepted Solutions (0)

Answers (2)

Answers (2)

Former Member
0 Kudos

Hi Ji,

I take it you have eliminated the possibility that the Repository host server time is not different to your pc, i.e. you think the job should be running but the according to the repository host server the start time has not arrived? The process that detemines when a job should be executed is the Dispatcher. If you look at the processes under each scheduler you'll see on one of the process Servers there is a dispatcher process. Is this process currently running on a machine that is very busy perhaps? Each Process Server has a dispatcher weighting parameter to enable you to influence where the Dispatcher process is executed, e.g. the Repository host server, most powerful machine, least busy machine etc.

Also take a look in the online help for the CPS registry parameter 'MaxAllocatedJobs'. Are you running lots of jobs concurrently which perhaps are being impacted by the default setting?

Regards,

Simon

Former Member
0 Kudos

Hi Simon,

Thanks for your indication. You might be right about this parameter: there is a lot of jobs running on the same process server, so that can explain the scheduled status.

The thing is now that by default no value is setted to this parameter in RW registry parameter. I assume that if there is no value, the value 10 is used. Am i right ?

Thanks for your help !

Regards,

Yi Jiang

Former Member
0 Kudos

Hi Yi,

yes, without a value set it would be defaulting to 10. The on-line help suggests a value in the range of 30-100 for a production environment. If you make a change start at the low end. The important thing to note here I think is that this does not affect the number of 'Running' jobs only the ability to initiate a bunch of jobs concurrently, i.e. my understanding would be that if you had say 50 jobs all scheduled to start at the same time by default the Dispatcher is only going to pick 10 in the first instance and only when each Process Server has confirmed that the job is running does the Dispatcher pick another etc.

What I am unsure about is, given that you say you can right click on a scheduled job and choose to execute immediatey, and it does, is whether this effectively overrides the MaxAllocatedJobs setting or whether in fact this registry setting is not the problem at all. I guess you could make the change and see what impact if any it makes?

Regards,

Simon

Former Member
0 Kudos

Hi Simon !

I have made the setting change (put the value 30 instead of 10). But this behavior still appears, and with the "Run job now" command, I can make the job run straight away...

So I think it's not related to this option.

Is there a way to monitor the dispatcher activities ? or anything that can make this process clearer ?

Thanks for your help!

Regards,

Yi Jiang

Former Member
0 Kudos

Yi Jiang,

Have you double-checked whether there are any timezone inconsistencies with these jobs? Check the submission details, queues, etc. to ensure they are all in the same timezone so that the jobs trigger when expected.

Regards

Guy

Former Member
0 Kudos

Hello Guy,

Thanks, but all the jobs are submitted in the same timezone. So I don't think that timezones are involved here.

Regards,

Yi Jiang