IBM Mainframe Forum

Hello experts,

Due to incorrect logic in programming, cics transactions are going into loop. what are the options to quickly identify such transactions to kill(these are not using more CPU) and prevent the damage for other systems which are impacting due to this looping transactions.

Thanks in advance.

So transactions are looping "due to incorrect logic in programming"...

Nobody in your organisation ever heard of walk-throughs, unit-testing, user-testing or acceptance-testing?

How dare you ask for help when all of the above would have identified this problem at about every earlier stage?

Are the looping transactions in a hard loop (no CICS commands) or soft loop (executing CICS commands)? Does your site have a CICS monitor such as Omegamon or MAINVIEW installed? If so, your best tool to identify the problem transactions would be the monitor tool -- by far. Otherwise, if there's a hard loop involved you may wind up having to take a dump to analyze; with a soft loop you can usually use CEDF to identify which transaction(s) hang around (any transaction showing up more than one or two hits of the enter key on CEDF is suspicious). Depending on the site and the looping involved, you may have no choice but to impact other systems (worst case: an IPL might be required).

In ALL cases, however, the LAST place you should be asking is a forum such as this -- we don't know your site, its tools, or anything else. You should FIRST go to your site support group for assistance.