Setting up auxiliary trace data sets
CICS® trace entries can be directed to auxiliary trace data sets, which are CICS-owned BSAM data sets. If you want to use auxiliary trace, you must create the data sets before you start CICS; you cannot define them while CICS is running.
About this task
- The internal trace table, in main storage in the CICS address space.
- The auxiliary trace data sets, defined as BSAM data sets on disk or tape.
- The z/OS® generalized trace facility (GTF) data sets.
- The JVM server trace file in z/OS Unix System Services
- You can collect large amounts of trace data, if you initially define large enough trace data sets. For example, you might want to do this to trace system activity over a long period of time, perhaps to solve an unpredictable storage violation problem.
- You can use the auxiliary switch (AUXTRSW system initialization parameter) to specify that auxiliary trace data cannot be overwritten by subsequent trace data. With the internal trace table, when the end of the table is reached subsequent entries overwrite those at the start of the table, and diagnostic information is lost.
- Auxiliary trace can be particularly useful if you are using CICS trace during startup, because of the high volume of trace entries that are written when CICS is initializing.
Auxiliary trace is held in one or two sequential data sets, on either disk or tape. The DD names of the auxiliary trace data sets are defined by CICS as DFHAUXT and DFHBUXT. If you define a single data set only, its DD name must be DFHAUXT.
Trace entries are of variable length, but the physical record length (block size) of the data that is written to the auxiliary trace data sets is fixed at 4096 bytes. As a rough guide, each block contains an average of 40 trace entries, although the actual number of trace entries depends on the processing that is being performed.