Common symptoms and actions

Use Table 1 to diagnose and correct problems.

Table 1. VIT analysis tool problems: Common symptoms and actions
Symptom	Action
Runs too long	If the tool is taking a long time to run (several hours): If the MATCH option was specified for storage analysis: Remove the MATCH option. Specify only the few pools you are interested in when using the MATCH option. Check to see if the tool is waiting on a tape to be mounted or to access a data set in use by some other job.
No output	If no output is displayed (for example, no matching VIT entries are found, no RUs are counted, no storage VIT entries are found): Verify that the trace has VIT records (see the SUMMARY data set). Verify that the entries required for the job are on the VIT. (All occurrences of VIT entries are listed at the bottom of the SUMMARY data set.) Storage analysis requires the SMS VIT entries. RU counting requires the PIU VIT entries. VIT extraction origin and destination options work only on PIU VIT entries. Check to see whether the VIT is a different level from the VIT analysis tool. For example, the DISP entry is now called DSP. Therefore, if you are extracting all occurrences of VIT option PSS with entry name e'DI*', no matches are found. Do not specify a start or stop time. An event reported on the console can be off several seconds from the GTF time stamp. Use the INTERVAL option to ensure that some output is seen before the job abends or is canceled.
Same output from a previous date	If the job runs but the output data sets contain data from a previous job, check the DISP parameter. When DISP is NEW and the data set exists, the batch job runs anyway and then deletes the new data set. A message in the JCL log indicates whether this has happened.
ABEND 80A	If you are running the storage analysis function with MATCH and LENGTH options, try one or more of the following methods: Increase storage on the job. Reduce storage pools to one storage type (GBLK, REQS, or VTAL). Match only a few pools rather than all GBLK or all VTAL or all REQS pools. Run the LENGTH option without the MATCH option, and save the output for future reference. The LENGTH option is independent of the MATCH option, so the same LENGTH output is shown, regardless of whether the MATCH option is specified. Remove MATCH and LENGTH options. Specify a start and stop time to limit the amount of data being processed.
Message CANNOT READ FILE WITH DD NAME TRACE received	When processing multiple tapes using the VIT analysis tools, you receive the message CANNOT READ FILE WITH DD NAME TRACE and the return code is 10. If you are attempting to process multiple standard label (SL) tapes using the bypass label process (BLP), verify that the LABEL parameter on the TRACE DD statement is coded correctly. See Table 1 to determine what document describes job control language (JCL).

In addition to the actions suggested in Table 1, try the following actions to help you diagnose the problem:

Use the DEBUG option (add it as a keyword in the parameter data set), which produces large quantities of data showing what the tool is doing. Run this on a small portion of the trace to prevent the output from being too large to be useful.
Run another tool, such as the IPCS GTFTRACE subcommand, or ACF/TAP to see whether they work on this trace data set and to compare output such as time stamps.
Run a short job to see what is on the tape. A simple way to run a short job is to run VIT extraction with an expression that is never true. Use the NOFORMAT option to avoid the overhead of loading the format routine. You can use the following parameter data set as a short job:
```
Desc Tell what's in this trace data set by running without extracting
Desc any entries.
NOWRAP NOFORMAT
VITEXT e'zzzz'
```
This data set extracts all VIT entries with the name ZZZZ. (Presumably, no entries start with ZZZZ.) The SUMMARY data set shows whether the data set wrapped, which types of records are traced, which VIT entries and options are traced, and the first and last time stamps.