Stack trace in joblog

The lob log has been for ages been the best way to find post mortem issues with applications. However the job log was designed back in time where the call stack was typical not huge.

Applications of today have huge call stacks and goes through various environments and when exception occurs very deep in the stack, it is not easy to find reasons for exceptions since the complete call stack is not exposed to the job when exceptions occurs.

This idea is about having the opportunity to set your job into a certain state so when an exception occurs then stack trace could be logged together with all the traditional job jog information.

If that is a separate new job log file or the existing job jog is not important as long is is possible to observe and correlate the call stack in a certain point of time for job in the current job log .

It could be made active by extending the job description - not only having the value of *SECLVL to perhaps *STKTRC for stack trace.

An option could also be to put the job in debug mode, however that will interfere with the same feature used in SQL so the job log will be polluted by too much information, so I would prefer not to do so.

This feature will improve traceability of interaction between ILE and PASE. it will improve traceability of stored procedures and user defined table functions and SQL in general.