[z/OS]

Extra inputs needed when reporting z/OS problems

[z/OS]When you report an IBM® MQ for z/OS® problem to IBM Support, you need to provide additional inputs to those that are needed for other platforms.

The core documentation that is required for reporting an IBM MQ problem on any platform is described in Contacting IBM Software Support.

As a rule, the total documentation you need to submit for a problem includes all the material you need yourself to do problem determination. Some of the documentation is common to all IBM MQ for z/OS problems, and some is specific to particular types of problem.

Make sure that the problem you have described can be seen in the documentation you send. If the problem has ambiguous symptoms, you need to reveal the sequence of events leading to the failure. Tracing is valuable in this respect but you need to provide details that trace cannot give. You are encouraged to annotate your documentation, if your annotation is legible and does not cover up vital information. You can highlight any data in hardcopy you send, using transparent highlighting markers. You can also write notes in the margins, preferably using a red pen so that the notes are not overlooked.

Finally, note that if you send too little documentation, or if it is unreadable, the support team will have to return your problem marked "insufficient documentation". It is, therefore, worthwhile preparing your documentation carefully and sending everything relevant to the problem.

The additional documentation needed is described in the next section. However, these are only guidelines. You must find out from the IBM support center representative precisely what documentation you need to send for your specific problem.

Additional documentation needed for problems with IBM MQ for z/OS

Here is a list of the additional documentation you might be asked to submit:
  • Any hardcopy or softcopy illustrating the symptoms of the problem.
  • The dump of the problem, see Getting a dump.
  • The appropriate SYS1.LOGREC records, see SYS1.LOGREC information.
  • The system log.
  • A portion of the IBM MQ for z/OS job log.
  • Trace records, see Using trace for problem determination on z/OS.
  • Trace information produced by the CICS® or IMS adapter, see Other types of trace.
  • Buffer pool statistics, see the sections on using SMF and type 115 records in Configuring z/OS .
  • Listings of relevant application programs.
  • A list of PTFs and APARs applied.
    Use the following sample JCL produce a list of all the PTFs, APARs, user modifications, and product code that has been installed in the global zone specified by SMPCSI:
    
    //SMPELIST EXEC PGM=GIMSMP,REGION=4096K
    //SMPCSI  DD DSN=shlqual.global.csi,DISP=SHR
    //SYSPRINT DD SYSOUT=*
    //SMPCNTL  DD *
    SET BOUNDARY(GLOBAL) .
    LIST
    SYSMODS
    APARS
    FUNCTIONS
    PTFS
    USERMODS
    ALLZONES .
    /*
    
  • Dump of coupling facility administration structure, see Figure 5.
  • Dump of coupling facility application structure, see Figure 5.
  • Dump of other queue managers in queue-sharing group, see Figure 1.
  • Definitions of the objects in your system.
    You can find these by using the DISPLAY command for each type of object, for example:
    
    DISPLAY QUEUE (*) ALL
    
    Or, if you regularly make a backup of your object using the MAKEDEF feature of CSQUTIL COMMAND function, the output from that backup job.
  • Your IBM MQ Db2 tables.

    You can get these by using the sample job CSQ45STB in thlqual.SCSQPROCS to produce a report of all the Db2 tables used by IBM MQ.

You might also be asked for the IBM MQ for z/OS symptom string, or for any keywords associated with the problem. The primary keywords are ABEND, WAIT, LOOP, PERFM, INCORROUT, MSG, and DOC, corresponding exactly with the problem classification types used in Building a keyword string. Strings containing other keywords are also useful. These are not predefined, and might include such items as a message or message number, an abend code, any parameters known to be associated with the problem, or, for example, STARTUP or INITIALIZATION. The keywords are then used by IBM Support as search arguments, to see if your problem is a known one.

Because of the size of SVC dumps in the cross memory environment, you might want to transfer the SYS1.DUMPxx data set to a tape or like device. You can use the PRDMP service aid program to transfer the SYS1.DUMPxx data set contents to another data set for archiving until the problem is resolved. Alternatively, your support center representative might give you the address of an FTP site where you can send your dump electronically. Each tape should be sent with the following associated information:
  • The PMR number assigned by IBM
  • A list of data sets on the tape (application source program, JCL, or data)
  • A list of how the tape was made, including:
    • The exact JCL listing or the commands used
    • The recording mode and density
    • Tape labeling
    • The record format, logical record length, and block size used for each data set
When the support team receives the package, this is noted on your PMR record on the RETAIN system.