A strategy for storage violations affecting innocent transactions
The storage violation has been caused by a program writing to an area it does not own, but you probably have no idea at the outset which program is at fault.
Look carefully at the content of the overlay before you do any other investigation, because it could help you to identify the transaction, program, or routine that caused the error. If it does not provide the clue you need, your strategy should be to use CICS tracing to collect a history of all the activities that reference the affected area.
The trace table must go back as far as task attach of the program causing the overlay, because that trace entry relates the transaction’s identity to the unit of work number used on subsequent entries. This could mean that a very large trace table is needed. Internal trace is not suitable, because it wraps when it is full and it then overwrites important trace entries.
Auxiliary trace is a suitable destination for recording long periods of system activity, because it is possible to specify very large auxiliary trace data sets, and they do not wrap when they are full.
If you have no idea which transaction is causing the overlay, you need to trace the activities of every transaction. This impacts performance, because of the processing overhead.