Issues Fixed in Cloudera Navigator 6.3.0

Navigator Audit Server sorts data when querying the backing database for audit information. When the backing database is Oracle, each query uses temporary tablespaces, which is cleared only when Navigator closes connections to the database. Over time Navigator queries can use up all of the temporary tablespace causing queries to fail.

This release includes a new configuration property maxConnectionAge that is set to one day (86400 seconds). If you continue to see that Navigator Audit Server is holding temp tablespace for too long for your Oracle database resources, you can adjust the setting to have Navigator Audit Server release the connections more frequently. Change the value in the Navigator Audit Server Advanced Configuration Snippet (Safety Valve) for cloudera-navigator.properties as follows:

navigator.max.connection.age=number of seconds

Note that if you have used Java Configuration Options for Navigator Audit Server to set the value of maxConnectionAge, that setting will no longer work.

This setting applies only to Navigator Audit Server with an Oracle backing database; it does not apply to other database types.

Cloudera Issue: NAV-7159

When streaming audit messages to both Flume and Kafka, if the Flume client throws an exception, Navigator Audit Server does not send the same messages to Kafka. To recover from this problem, the Flume client needs to be working.

Cloudera Issue: NAV-7143

Navigator Metadata Server version 6.2 was unable to start if the host where it is installed did not have access to the Internet. This problem is fixed in this release.

Cloudera issue: NAV-7115

The commons-fileupload package used by Navigator has been upgraded to 1.3.3 to resolve a security vulnerability indicated by CVE-2016-3092.

Cloudera issue: NAV-7099

Navigator support for Compute-only clusters in Cloudera Manager includes audit from services on the Base cluster and Compute clusters. No metadata is extracted from services running on the Compute cluster.

For more information, see Auditing in Virtual Private Clusters.

Cloudera Issue: NAV-7091

Navigator Metadata Server purge could produce messages such as "Checking if maintenance is running" and then failed to run during the available window. This problem occurred when a scheduled purge job waits for extraction tasks to finish but while waiting, a policy job starts, preventing the purge job from running. In this release, policy jobs are not be triggered if purge jobs are scheduled and waiting to start.

Cloudera issue: NAV-7037

Under some circumstances, Navigator Metadata Server created invalid operation to operation execution relations. This caused errors when displaying the lineage in the Navigator console. Also, the metadata purge could not delete the invalid relations. This problem is fixed in this release.

Cloudera Issue: NAV-7035

Navigator tags now support UNICODE characters beyond ASCII. Only ASCII text can be used in the name of a user-defined or managed property. Property values can include international characters.

Cloudera Issue: NAV-7011

This release includes changes that improve how Navigator Metadata Server uses memory, reducing the overall heap required.

Cloudera Issue: NAV-6958

If a truststore password is not set, Navigator Metadata Server would not start. Cloudera Manager showed the status of "Role failed to start due to error null." This problem is resolved in this release: if TLS is enabled, Navigator Metadata Server has the same behavior as other Cloudera Manager-managed services and will start without a truststore password.

Cloudera issue: NAV-6829

When an error occurred in the audit pipeline, it was possible that not all components in the Navigator Audit Server recovered fully. One symptom of such a failure is the "CannotSendRequest" error in the Navigator Audit Server log. This problem is fixed in this release.

Cloudera issue: NAV-6321