CICS® monitoring facility enables you to collect the following types, or classes, of monitoring data:
Performance class data is detailed transaction-level information, such as the processor and elapsed time for a transaction, or the time spent waiting for I/O. At least one performance record is written for each transaction that is being monitored.
Performance class data provides detailed, resource-level data that can be used for accounting, performance analysis, and capacity planning. This data contains information relating to individual task resource usage, and is completed for each task when the task terminates.
You can enable performance-class monitoring by coding MNPER=ON (together with MN=ON) in the system initialization table (SIT). Alternatively you can use either the (CEMT SET MONITOR ON PERF) or EXEC CICS SET MONITOR STATUS(ON) PERFCLASS(PERF) commands.
This information could be used periodically to calculate the charges applicable to different tasks. If you want to set up algorithms for charging users for resources used by them, you could use this class of data collection to update the charging information in your organization’s accounting programs. (For older versions of CICS, we did not recommend charging primarily on exact resource usage, because of the overheads involved in getting these figures.)
Exception class monitoring data is information on CICS resource shortages that are suffered by a transaction. This data highlights possible problems in CICS system operation and is intended to help you identify system constraints that affect the performance of your transactions. There is one exception record for each type of exception condition. The exception records are produced and written to SMF as soon as the resource shortage encountered by the transaction has been resolved. Exception records are produced for each of the following resource shortages:
If the monitoring performance class is also being recorded, the performance class record for the transaction includes the total elapsed time the transaction was delayed by a CICS system resource shortage. This is measured by the exception class and the number of exceptions encountered by the transaction. The exception class records can be linked to the performance class records either by the transaction sequence number or by the network unit-of-work id. For more information on the exception class records, see Exception class data.
You can enable exception-class monitoring by coding the MNEXC=ON (together with MN=ON) system initialization parameters. Alternatively, you can use either the CEMT command. (CEMT SET MONITOR ON EXCEPT) or EXEC CICS SET MONITOR STATUS(ON) EXCEPTCLASS(EXCEPT).
Transaction resource class data provides additional transaction-level information about individual resources accessed by a transaction. Currently, the transaction resource class covers file and temporary storage queue resources. The maximum number of files and temporary storage queues monitored for each transaction is limited by the FILE and TSQUEUE parameters on the DFHMCT TYPE=INITIAL macro. The default is FILE=8 for files and TSQUEUE=8 for temporary storage queues. Therefore, you may need to assemble an MCT that specifies either or both FILE and TSQUEUE options if the default values are insufficient, or if you do not want to collect transaction resource data for either files or temporary storage queues. One transaction resource record is written for each transaction that is being monitored, provided the transaction accesses at least one of the resources for which monitoring data is requested, (for example, at least 1 file if you specify FILE=number).
Performance class data also provides information about file and temporary storage queue resource accesses, but this information in the performance record is given in total only for all files (see Performance data in group DFHFILE) and all temporary storage queues (see Performance data in group DFHTEMP). Transaction resource data breaks this information down by individual file name and temporary storage queue name, up to the maximum number specified in the MCT. Transaction resource information is completed for each task when the task terminates.
You enable transaction resource class monitoring at startup by coding MNRES=ON (together with MN=ON) as a system initialization parameter. Alternatively, you can use one of the following commands to enable performance class monitoring dynamically:
If you are running CICS with the MVS™ workload manager in goal mode, the MVS workload manager provides transaction activity reporting by service class and/or report class, based on transaction response time information.
See MVS workload manager for more information about the MVS workload manager.