Troubleshooting › Identify and Resolve the Problem › Event Management Troubleshooting Checklist

Event Management Troubleshooting Checklist

The following items may help you identify the problem.

1. Are z/OS UNIX Services configured and operational?

   You should be able to initiate a USS shell session and issue various UNIX commands to verify that z/OS UNIX Services are operating correctly. Pay particular attention to the USS configuration parameters specified in the Installation Guide.

2. Are proper security definitions in place?

   Event Management components must run with superuser privileges (UID 0). Verify that caiopr and stardaemon both start and run with UID 0. If you have activated BPX.DAEMON support, the user ID running Event Management needs to be permitted to the BPX.DAEMON and BPX.SERVER FACILITY resources.

   In addition, depending on the z/OS release you have installed, this ID may also require both SURROGAT and BPX.SERVER.userid permissions for any users that run Event Management commands.

3. Is a compatible Java environment installed?

   The Event Management Java GUI requires Java at a JDK 1.1.6, 1.1.8, 1.3.1 or 1.4 level (available from IBM). Verify that simple Java programs (such as the JDK supplied samples) compile and execute properly.

4. Is your web server configured and running properly?

   You should be able to request a simple HTML document by connecting a web browser to the mainframe and requesting one of the CA Common Services for z/OS HTML documents. If, for example, your host name is "mainframes.com" and you have selected port 4080 to run the Services web server, then this URL should display the CA Common Services for z/OS welcome page:

   http://mainframes.com:4080/tngfw/tngfw.html
   

5. Are all of the required processes running?

   From a UNIX shell session or batch job with superuser privileges, issue "ps -ef" (or the z/OS command "D OMVS,A=ALL") and verify that these processes are running:

   • Event Management opr component requires caiopr, cailgr and newdaylog processes. Optionally, if store and forward is required, an oprsafd processing should be running.
   • Star component requires stardaemon process
   • Calendar component requires calendar process
   • Java running class CA.Unicenter.comm.w2RmiServerImp
   • logonserver.exe
   • w2Tree.exe
   • CaemRtS
   • EMServer.exe
6. Are your environment variable settings correct?

   There are a number of environment variables that must be correctly set in order for Event Management to operate correctly. These environment variables are set in the $CAIGLBL0000/PROFILE file for the Event Management processes and in the $CAIGLBL0000/browser/httpd.envvars file for the web server and Java GUI processes.

   Tracing is usually turned on by setting environment variables before the process is started. The output is normally directed to STDOUT. The following variables are used to start tracing for caiopr. Additional environment variables for tracing other specific processing can be set at the direction of CA Support.

   CA_CAIDEBUG=Y
   CAI_NODENAME_DEBUG=Y
   

   Some trace messages will be written based on the existence of file /cai/nsmem/emsrvc/config/traceinfo. That file contains control statements that direct the trace to a file, syslogd, or both. Send them to a file because it is easier to send the file for analysis.

   The trace file that is written to is /cai/nsmem/emsrvc/config/debug.log.

   Note: For more information, see this guide and the Reference Guide.

7. Are the CA Datacom/AD libraries in your STEPLIB concatenation?

   z/OS UNIX Services processes require the CA Datacom/AD CUSLIB and CAAXLOAD data sets to be defined in the STEPLIB environment variable.

8. If you are having trouble connecting Event Management to other platforms, you may need to configure CAICCI on z/OS and on the remote system. For more information on defining remote nodes to the z/OS system, see the Common Communications Interface chapter.
9. For more information about defining the z/OS system to the remote distributed machine, see the CA NSM documentation.
10. If you are having problems logging on to the Java GUI from your web browser, verify that the program logonserver.exe in directory $CAIGLBL0000/wv/bin is running APF authorized and program controlled.

    You can verify this with the following USS command:

    extattr $CAIGLBL0000/wv/bin/logonserver.exe
    

11. If you are having trouble capturing syslogd messages, verify that the syslogd daemon is running, and that the syslogd configuration file correctly forwards messages to Event Management. This can be checked by ensuring that the /etc/syslog.conf file contains a line for the node's pipe file such as:

     .info.../cai/nsmem/opr/config/USCMCT3Q/pipe/oprpipe0001
    

12. If you are having trouble running CATRAPD, make sure that SNMP port 162 is available and can be opened by CATRAPD. Your TCP/IP configuration file can prevent certain ports (such as 162) from being opened except by the specified application.

    By default, IBM ships TCP/IP pre configured so that port 162 can be opened only by a program called SNMPQE, and this must be removed before CATRAPD can operate correctly. Alternately, you can specify a different port by editing and exporting an environment variable. For more information, see the Installation Guide.