Esteemed listers, I have been tasked to put a procedure in place that will generate an alert when a batch job "takes longer than usual" to complete. Yes, I know, that is a rather vaguely stated requirement...
I know I can find figures about the elapsed time of a batch job in one of the SMF records. I think I can build a database of historic runtimes and with a bit of thought, I am sure I can come up with a query that will list me the "outliers". But in order to generate an alert when a job takes too long, I will need some mechanism that gives me the information about that job while it is still running. And while I could perhaps interrogate the system (although I don't have a clear idea on how to actually do that) every so often about the running jobs and compare the time taken of each running job to the database, I am afraid that would take far too much resources. Indeed, with several hundreds if not thousands jobs running at any give time, the task of getting all the figures of these running jobs and comparing this to the database will definitely not be cheap in system resources. We don't have any budget for vendor products, I am afraid... And we're short of MSU's... Any suggestions? Thanks and very best regards, Jantje. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
