Successor jobs

[pavan shashikanth guddeti]

Hi All,

We are facing a Scenario where if my predecessor Job gets Fault impossible start Error then my Successor job is getting Ended inactive and for my Successor jobs i do have Conditions in my time and Dependencies status check as Any_ok (Number 4 as per Screenshot)and Ended_Not_ok(Number 5 as per Screenshot).Below is the Screenshot can i know if i needs to make any changes to my Successor job Status check ? 
And we do have close to 1000 jobs in same way is there any other process where my Successor job will show as waiting for Predecessor as a regular failure.

[Pete Wirfs]

What is your use-case for "FAULT_OTHER - start impossible." to be tolerated?  We don't tolerate this level of failure.  It is catastrophic.  It would be better if you would modify POSTLOAD.CUBE so it can shut down gracefully (i.e. skip) when it encounters the issue that wants to throw "start impossible.".

------------------------------
Pete Wirfs
SAIF Corporation
Salem Oregon USA
------------------------------

[pavan shashikanth guddeti]

Thanks pete for the quick response.We require if POSTLOAD.CUBE gets FAULT_OTHER - start impossible then BI.LOG.END.SUCCESS and BI.LOG.END.FAILURE should Show as Waiting for predecessor(just as how One Automation Successor jobs waits for the if predecessor job fails).

And we are having around ~1000 workflows which are having same structure/design .So, if possible can we have a way where we can make changes in process tab of the POSTLOAD.CUBE which will take care.

[Pete Wirfs]

Do you know what causes FAULT_OTHER - start impossible? 


------------------------------
Pete Wirfs
SAIF Corporation
Salem Oregon USA
------------------------------

[pavan shashikanth guddeti]

HI Pete,

We aware of some scenarios like
1)Error code '1314', error description: 'A required privilege is not held by the client.'
2)if we dont have the calendar event 

[Pete Wirfs]

I think the catastrophic failure of FAULT_OTHER invalidates the entire workflow, in which case what you are trying to accomplish would be impossible.

My advice is to continue looking for ways to avoid  FAULT_OTHER.  Maybe engineer a task that runs before POSTLOAD.CUBE to test the viability of POSTLOAD.CUBE and tell it to skip if it is not viable?

We generate daily cubes here, but with very few workflows to maintain.  We run a script that kicks off multiple instances of our model workflows.




------------------------------
Pete Wirfs
SAIF Corporation
Salem Oregon USA
------------------------------