IM InfoSphere Identity Insight, Version 8.0

Pipelines troubleshooting checklist

If you experience problems with the pipelines, before calling IBM Software Support, review this list of the most common pipeline problems encountered.

  1. Pipeline reports a status of Down or is not reporting any status
  2. Pipeline shuts down
  3. Pipelines do not honor configuration changes made in the Configuration Console
  4. Pipelines do not start on AIX
  5. Pipeline processes only part of an incoming record
  6. Transport does not work
  7. Pipeline does not load scientific notations or floating point numbers
  8. After starting a pipeline, I receive a warning message that indicates no routes are defined
  1. Pipeline reports a status of "Down" or is not reporting any status at all
    • If you are looking at the Pipeline Status tab in the Configuration Console, consider the following questions:
      • Is the pipeline registered on the Pipelines tab?
        • If you do not see the pipeline name on the Pipelines tab, this pipeline is not registered and does not display status on the Pipeline Status tab. Register the pipeline.
      • Was the pipeline started using the same name as its registered name on the Pipelines tab? Remember that the pipeline name typed on the command line must be an exact match, including case, to the registered name.
    • Is the SNMP agent running on the pipeline node?
      • If not, start the SNMP agent. Make sure to specify the port_number using the -p option.
    • Is your SNMP agent using the correct port number?
      • If your system was installed using this default SNMP agent port number (13516) and your system uses a different port number, use the configuration utility to edit the port number.
      • If you do not remember which port number was configured at the time of installation, check the server.xml file, located on each pipeline node in the directory where the SNMP agent is installed.
    • Does the pipeline node have an error or is it not running?
    • Has the name of the pipeline node changed from the pipeline node registered to this pipeline name on the Pipelines tab in the Configuration Console?
    • Did the transport specified in the pipeline command use the correct syntax?
    • Has the pipeline shut down?
  2. Pipeline shuts down or crashes
    • Did the pipeline encounter too many errors processing incoming data files?
      • Check the log files for more information about the errors. Use that information to resolve the problem.
      • Check the ErrorLimit setting in the pipeline configuration file. You might need to increase this number.
    • Did the pipeline run out of memory resources?
    • Is the database causing the problem because of one of the following reasons:
      • Not enough disc space?
      • Lost connectivity to the pipeline?
      • Has the user name and password changed for this database?
  3. Pipelines do not honor configuration changes made in the Configuration Console
    • Before pipelines apply configuration changes, they must be stopped and restarted. When the pipelines restart, the configuration changes are applied as part of the pipeline initialization process.
    • For data-integrity purposes, stop and restart all running pipelines after a configuration change.
  4. Pipeline does not start on AIX
    • Did you receive an error message indicating that the "dependent module libicuio.a could not be found"?
      • If so, make sure that the library is found in one of the following directories: /usr/lib, /lib, $DB2INSTHOM/sqllib/lib. Or set the LIBPATH environment variable to include the product installation_home/lib directory.
    • Check the version and location of the C++ runtime libraries. The problem might be a result of incorrect settings in the RunTime Update and the LIBPATH environment. Please refer to the "IBM InfoSphere Identity Insight Installation and Configuration Guide" or to the latest support information.
  5. Pipeline processes only part of an incoming record and not the full record
    • Check the *.BAD log file for invalid UMF messages. This log file indicates the name of the incoming data source file that was being processed.
    • Check the UMF Exceptions tab in the Configuration Console.
    • Generate the Data Source Summary report or the Load Summary report and look for poor quality data. Review the affected records in their associated data source files. When you have corrected the problem, reload the corrected records into a pipeline for processing.
  6. Pipeline transport does not work
    • Make sure that the syntax used for the transport is correct. For example, if you are specifying a database transport, did you include quotes in the appropriate place?
    • If the transport is a queue transport, does the message queue exist?
    • If the transport is a file, does the file exist? Is the file located in the directory specified in the transport?
    • If the transport is a database transport, is the platform IBM DB2® on z/OS®? If so, the database transport is not currently supported on this platform.
  7. Pipeline does not load a scientific notation or a floating point number
    • This is a known pipeline limitation. Revise the scientific notations or floating point numbers in the UMF to multiply out the exponent, so the number is in standard numeric notation. For example, -1.267E-05 multiplied out is -0.00001267.
  8. After starting a pipeline, I receive a warning message that indicates no routes are defined
    • This is an information warning message only. It is safe to ignore it. (The message just informs you that there are no routes defined for this pipeline. Routes are not required to run a pipeline.)


Feedback

Last updated: 2009