Troubleshooting and Limitations

This section includes the following:

Limitations

  • The Data Flow Probe or Integration Service must be installed on a Windows OS.

  • For requests created in APM from CIs pushed from UCMDB, any changes made in APM are overwritten when you run the data push job in UCMDB.

  • The APM Application request form holds a single value for Location, therefore it is designed to push only one value for Location of the Application from the UCMDB BusinessApplication.

    In the definition of Location of the Application (see the pushMappingAPMApplicationR.xml file), the Location of the Server is used for the Location of the Application. For an Application that contains multiple servers, select one of the servers and then you can get its Location.

  • For value mappings between UCMDB and APM, certain mapping rules are followed.

    For example, when synchronizing the Business Criticality field of Application, the following mapping rule is used for the mapping:

    <target_mapping name="REQD.APM_RATING_BUSINESS_CRIT" datatype="STRING" value="APMPushFunctions.getPropertyValue('bc', Root['business_criticality'].toString() , '')"/>

    where the definition of ‘bc’ token is defined as follows in the <APMPushAdapter.zip>/mappings/scripts/server.properties file:

    bc.0=0 - Least critical
    bc.1=1 - Slightly critical
    bc.2=2 - Less than average
    bc.3=3 - More than average
    bc.4=4 - Critical
    bc.5=5 - Highly critical

    When synchronizing the Business Criticality field from UCMDB to APM, if the value is ‘1’ in UCMDB, then the field value will be set to ‘1 - Slightly critical’ in APM.

    If you need to use this value mapping for other fields from UCMDB to APM, make sure you customize the mapping by following the example above.

  • This integration does not support synchronizing values in languages that are not supported by UCMDB. For example, Simplified Chinese.
  • The ";" character is not supported. If a UCMDB CI name contains a ";" character, it would be treated as a separator and two entries would show up in APM after synchronization.

Troubleshooting Problems

  • Problem: Some UCMDB CIs include characters that are not supported in APM Entities. For example, semicolon (;).

    Solution: The suggested solution is to modify the content synchronized from UCMDB to APM. You can use the Replace function to replace the unsupported characters for this field mapping in the XML mapping file. However, note that this may cause inconsistent content between UCMDB and APM.

    An example, in the pushMappingAPMApplication.xml file, the semicolon character (;) is replaced with a space:

    <target_mapping name="REQ.KNTA_PROJECT_NAME" datatype="STRING" value="APMPushFunctions.stringReplace(Root['name'], ';', ' ')"/>

  • Problem: For some fields, the field value lengths between APM and UCMDB are different, therefore you may need to customize the field mapping. You can follow the example below:

    Example

    Use a substring as illustrated below to limit the field length to 200 characters:

    <target_mapping name="REQ.DESCRIPTION" datatype="STRING" value="APMPushFunctions.subString(APMPushFunctions.stringReplace(Root['name'],';',' '),200)"/>

Logs

The push adapter framework uses a different logging system for the normal fcmdb.adapters.*.log files.

To change the level of the log files to debug, edit the following file:

  • On the Data Flow Probe machine:

    ..\DataFlowProbe\conf\log\fcmdb.push.properties

  • If using the integration service, on the UCMDB server:

    ..\UCMDB Server\Integrations\conf\log\fcmdb.push.properties

Change the log level to DEBUG:

loglevel=DEBUG

The integration generates fcmdb.push.* logs in the following folder:

  • On the Data Flow Probe machine:

    ..\DataFlowProbe\runtime\log\

  • If using the integration service, on the UCMDB server:

    ..\UCMDB Server\Integrations\runtime\log\