Troubleshooting – Universal Discovery

Problem. Data that should have been discovered during the discovery process does not appear in the topology map.

Verification. The CMDB cannot retrieve the data or build the query results. Check the Discovery Results pane. If the CIs were not created, the problem is occurring during the Discovery process.

Solution. Check the error messages in the probeMgr-services.log file located in C:\UCMDB\DataFlowProbe\runtime\logs.

Problem: There are triggers running in the Management Zone that should not be running.

Reason: Running triggers continue to run in the case where a Probe cluster is bound to the Management Zone, and the Probes in the cluster are removed from the cluster while the triggers are running.

Solution: To stop the triggers running, deactivate and then reactive the Management Zone.

Problem: A discovery job is running triggers that are not within its Probe limit.

Indication: Triggers are not released from a job in the following cases:

Scenario 1

1. ProbeA and ProbeB belong to Cluster1.
2. Range IPs by ICMP is limited to run only on Cluster1. The job runs on both Probes in Cluster1.
3. ProbeB is removed from Cluster1.
4. In the next scheduled run of Range IPs by ICMP, you notice that the trigger is still running on both Probes in the cluster, even though ProbeB no longer belongs to Cluster1.

Scenario 2

1. ProbeA is in Cluster1; ProbeB is under the Default Domain
2. Range IPs by ICMP is limited to run only on ProbeB.
3. ProbeB is added to Cluster1.
4. In the next scheduled run of Range IPs by ICMP, you notice that the trigger is still running on ProbeB, even though ProbeB now belongs to Cluster1.

Solution. Deactivate and then reactive the job.

Problem. Not all networks or IPs have been discovered.

Indication. Not all the networks or IPs appear in the topology map results.

Verification. The IP address range in the Data Flow Probe Setup window does not encompass the scope of the networks or IPs that should have been discovered.

Solution. Change the scope of the Discovery range:

1. Select Data Flow Management > Data Flow Probe Setup.

2. Select the Probe and the range.

3. Change the IP address range in the Ranges box as required.

Problem. Not all TCP ports have been discovered.

Indication. Not all TCP ports appear in the topology map results.

Verification. Open the portNumberToPortName.xml file (Data Flow Management > Adapter Management > DDM Infra > Configuration Files > portNumberToPortName.xml), and search for the missing TCP ports.

Solution. Add the port numbers that should be discovered to the portNumberToPortName.xml file.

Problem. Failure to discover resources on a machine running on the Windows platform.

• Solution 1. Start > Settings > Control Panel > System. In the Remote tab, verify that the following check box is selected: Allow users to connect remotely to this computer.

• Solution 2. In Windows Explorer, select Tools > Folder Options. In the View tab, clear the Use simple file sharing (Recommended) check box.

Problem. If you notice that trigger CIs for jobs in a management zone are in continuous "Progress" status, configure Data Flow Probe to ignore certain call home requests from the Universal Discovery Agent.

• Solution. To resolve this issue, change a parameter value as follows:

  • GlobalSettings.xml file
  • allowCallhomeInterval parameter
  • Default is 24.
  • Measured in hours that call home requests are ignored.
  • Allowable values are any integer greater than 0.

Problem. If you notice that certain devices contain unexpected values or contain no values.

• Solution. To resolve this issue, run the Rulebase Support Report and send to Support for analysis. For more information, see Universal CMDB Modeling Guide.

Problem. If you notice that the Inventory Discovery by Scanner job fails.

Indication. The Communication log contains the following entry: "Step Wait XML Enricher Process execution failed.".

Solution. If BDNA Normalize integration is enabled, troubleshoot BDNA Normalize operation using its documentation or contact BDNA support.

• How to get a view of the integrations

  To get a view of the integrations, run the following command:

  SELECT ds.datastore as integration_point_name ,COUNT(*) as nbr_of_records ,ds.ds_id FROM SYNC_ID_MAP as ID, SYNC_DATASTORE as DS where id.ds_id = ds.ds_id group by ds.datastore,ds.ds_id

• Clear a given datastore by invoking the removeIdMappingsOfDataStore JMX method

  To clear a given datastore, you can invoke the UCMDB:service=FCMDB Synchronizer Services > removeIdMappingsOfDataStore JMX method.

  Caution Make sure you do it only when necessary.

  For SM adapters, invoking the removeIdMappingsOfDataStore JMX method will resend all CIs, and then create duplicates.

  However, in general it is useful for AM adapter and Generic Adapter, when there is assumption that the data push is corrupted.

Problem. Test connection failed for UCMDB-UCMDB integration where one UCMDB uses SSL authentication for probes and the other does not.

Solution. The connection is built from the non-SSL authentication probe to the target UCMDB server that uses SSL authentication for probes. In this case, the target UCMDB server cert should be imported into the non-SSL probe truststore:

<DataFlowProbe>\conf\security\HPProbeTrustStore.jks (the SSL configuration is included in ssl.properties)

For details about how to import the cert, see Enable SSL with Server (One-Way) Authentication.