Interpreting the evidence

The first things to look at are any messages that accompany the abend, the exception trace entry in the internal trace table, and the symptom string at the start of the dump.

Procedure

  1. Look at any messages that accompany a CICS® system abend, because they might point directly to the cause of the failure. For advice about the user response to a message, see CICS messages.
  2. Examine the exception trace entry. The exception trace entry gives information about what was happening when the failure occurred, and data that was being used at the time. When a CICS system abend occurs, an exception trace entry is made to the internal trace table and any other active trace destination. It does not matter whether you have tracing turned on; the trace entry is still made.

    If the trace table contains more than one exception trace entry, it is likely that the last one is associated with the dump. However, this might not always be the case, and you should make sure that you have found the correct entry. Be aware, too, that dumps can sometimes be requested without a corresponding exception trace entry being made.

    For details of trace entries, see CICS trace.

  3. Look at the symptom string in the system dump. The symptom string, similar to the short symptom string at the beginning of a transaction dump, appears at the beginning of a CICS system dump. It is also written to SYS1.LOGREC and is issued as part of message DFHME0116.

    The symptom string provides a number of keywords that can be directly typed into RETAIN and used to search the RETAIN database. The possible keywords are shown in Table 1. The keywords are used at the IBM® Support Center to discover duplicate problems, or problems that have already been reported by other users and for which a solution is available.

    Table 1. Symptom string keywords
    Keyword Meaning
    PIDS/ Product ID (CICS product number)
    LVLS/ Level indicator (CICS release level)
    RIDS/ Module name
    PTFS/ Module PTF level
    MS/ Message ID reporting error
    AB/ Abend code
    ADRS/ Address or offset indicator
    PRCS/ Return code
    PCSS/ CICS jobname
    OVS/ Overlaid storage
    FLDS/ Name of a field associated with problem
    REGS/ Software register associated with problem
    VALU/ Value of a named field or register

    Although the symptom string is designed to provide keywords for searching the RETAIN database, it can also give you significant information about what was happening at the time the error occurred, and it might suggest an obvious cause or a likely area in which to start your investigation. Among other things, it might contain the abend code. If you have not already done so, look in CICS messages to see what action it suggests for this abend code.

    If the system cannot gather much information about the error, the symptom string is less specific. In such cases, it might not help you much with problem determination, and you need to look at other parts of the dump. The kernel domain storage summary is a good place to start.



dfhs10v.html | Timestamp icon Last updated: Monday, 5 June 2017