Terminal error recovery
The resolution of certain conditions (for example, permanent transmission errors) involves both CICS® and additional user coding. CICS cannot arbitrarily take all action with regard to these errors. User application logic is sometimes necessary to resolve the problem.
For the z/OS® Communications Server part of the network, terminal error handling is carried out by the node abnormal condition program (NACP) and a sample node error program (NEP), provided by CICS, or a user-written node error program. For further information about these, see Node abnormal condition program and Node error program.
For the portion of the telecommunication network connected to BSAM, these error-handling services are provided by the terminal abnormal condition program (TACP) and by the user-written or sample terminal error program (TEP). For further information about these, see Terminal abnormal condition program and Terminal error program.
- The terminal is “locked” against use.
- The node or terminal abnormal condition program is attached to the system to run as a separate CICS task.
- The node or terminal abnormal condition program writes the error data to a destination in transient data control if the user has defined one. This destination is defined by the user and can be intrapartition or extrapartition.
- The node or terminal abnormal condition program then links to the appropriate node/terminal error program to allow terminal- or transaction-oriented analysis of the error. In the node or terminal error program, the user may decide, for example, to have the terminal placed out of service, have the line placed in or out of service, or have the transaction in process on the terminal abnormally terminated.
- The terminal is “unlocked” for use.
- The node or terminal abnormal condition program is detached from the system if no other terminals are to be processed.