DSXchange

Hello friends,

We have a unique problem. We have a job A which is triggered by seqA and we use a triggering tool to trigger all our jobs

The problem is, job A is not getting triggered immediatley after the tool triggers seqA job. Somehow job A is waiting for few minutes before starting

Did anyone notice this before? Any help would be highly appreciated

Thanks a lot in advance
Yamini

Hi Yamini,
What is the desing of your JobSequence? Do you have any before Job subroutine or any scripts that been called. How many parallel jobs/JobSequence been called during the run? How busy was your server during the time of this jobsequence been triggered? How many stages in the Job that been called by Sequence?
If you have only one Jobsequence and only one job been called in that sequence, without any before job actions, it should start pretty much quick.

What is your "triggering tool"? There is usually some startup overhead, more in parallel jobs than in server jobs. There is also a delay before the Director client refreshes. What evidence do you have that job A is "waiting"? How do you measure this? Is ANY of the processes in job A (for example DSD.RUN) starting? What is the timestamp in the "job starting" event in the job log?

yaminids wrote:Somehow job A is waiting for few minutes before starting

There really isn't any way for the Sequence job to 'delay' starting JobA unless you have activities in the Sequence job before the Job Activity stage for JobA.

One situation I've seen that confuses people is if JobA has a 'before job' activity that it runs that takes 'a few minutes'. It will look like there is a delay starting the job when in fact it has started but the before job activity is running. That's one reason I insist on an Annotation on the canvas clearly stating the fact when Before or After job activities are afoot in a job. Any chance that is the case here?

You can do another comparison.
Run the JobA which is Job sequence outside the sequence. Does it take a long time to start?
What does JobA do? Are you running a intensive process, like querying a large database?
That could be a starting point for the resolution.

I suspect that the job is starting porperly, but the status is not updated promptly. Tell me this, if you run the job manually from director, do you observe the same delay?

Ray,

We use Espresso to trigger our jobs (about 100.) All jobs run without any issues except the one in question.
Job A is triggered sequentially by three different sequences passing different parameters

When I check Director logs there is a difference of atleast 5 mins between Sequence and the actual job

Any thoughts?
Yamini

Can you please post (a) the "job run requested" message from the job sequence's log, and (b) the "job starting" message from job A's job log?
Full and complete detail please.

Is it possible that Job A is already running from one sequence when you try to trigger it from another? If so, have you considered making Job A a multi-instance job and triggering it with different invocation IDs?

I job might have already started but the status still shows that its not running. Might be a delay in updating the RT_STATUSnnn file. Try clearing the status from the director, clearing the &PH& folder will also help in quicker response times.

A 5 minute delay in updating status or the log?

How about my 'thought' that seems to have been completely ignored? Did you check for a 'before job' activity in JobA? Get that card off the table if it isn't a factor.

Craig, even for a before-job activity, doesnt the job start first and then kick off the before job acitivity. It still shows the status as Running even though the before job activity is running.

Yes, it does - but what shows in the Monitor can be confusing and can make it look like 'nothing is running'.

BTW, the OP has never explained exactly what this means or how they are measuring this 'delay':

yaminids wrote:The problem is, job A is not getting triggered immediatley after the tool triggers seqA job. Somehow job A is waiting for few minutes before starting

Oh, sure, there's been the oh so typical chatter back and forth between us girls postulating like mad - but no clarification from the OP.

Ray,

The following is the information from the log

Sequence log:
Starting Job MMI002PSeq.
$PR_DSN = XXXXXXX
$PR_USERID = XXXXXXXX
$PR_PASSWORD = ********
$MM_DSN_ORA = XXXXXX [$PROJDEF]
$MM_DSN = XXXXXX [$PROJDEF]
$MM_USERID = [$PROJDEF]
$MM_PASSWORD = [$PROJDEF]
$NLS_LANG = AMERICAN_AMERICA.AL32UTF8 [$PROJDEF]
$DS_Q_AUDIT = AUDIT_IN [$PROJDEF]
$DS_Q_MGR = XXXXXX [$PROJDEF]
pAppId = MM
pSubAppId = PR
pIntfName = MMRI002
pRunId = 2
$MM_DSN_PL = RPTMCM [$PROJDEF]
$MM_USERID_PL = DS_MCMBGWIP [$PROJDEF]
$MM_PASSWORD_L = ********
pSRCDB = BM

Job log:
Starting Job MMRI002.
$PR_DSN = XXXXX
$PR_USERID = XXXXX
$PR_PASSWORD = ********
$MM_DSN = XXXXXXX
$MM_USERID XXXXXXX
$MM_PASSWORD = XXXXXX
$NLS_LANG = AMERICAN_AMERICA.AL32UTF8
pSRCDB = BM
DSJobController = MMI002PSeq

I have noticed that the job was not running when one of the sequences was trying to trigger it.

Craig,
Sorry for not gettting back to the post as I was away from work. Anyway, the job doesn't have any 'before job' activity

I am measuring the delay by comparing the start time of the Sequence to the start time of the job

Any thoughts friends?
Yamini

What are the timestamps on the two "starting job" log events?

What happens when you trigger the JobSeqA manually (without being scheduling). How many other jobs were triggered parallely by the Sequence?