Re: *** ADMIN: Wiki spam endjob\dltpgm

On 08-Sep-2011 11:36 , Dennis wrote:

I know this post was in jest, but for the record: that won't end a
job that has an active monitor for MCH* messages.

"Musselman, Paul"<pmusselman@xxxxxxxxxxxxxxxx> wrote:

Won't the "DLTPGM" command end a running job as well? (:

FWiW...

A "call" or "transfer control" to a program that has previously been deleted would effect an error condition that can be monitored. Though depending on the mechanism for the call request, not necessarily manifest to the caller as a message in thee MCH0000 range.

However AFaIK any "return" to a previous invocation in the call stack which is to a non-existent program is not a condition that can be monitored. In my experience, for that scenario the job ends with a function check VLog, manifest to the *EXT program message queue as CPC1220, even though MCH3203 is the diagnosed error code. Of course the job could possibly continue running fine up until that return instruction. In the lab I used to take advantage of that effect to resolve various "loop" scenarios that others would insist required a pwddwn\IPL to recover.

Regards, Chuck