Operations by the Fault Monitor During a Probe

The probing consists of checking to see if kadmind(1M) and krb5kdc(1M) are listening to their respective ports. During more thorough probing a new principal is created, the principal is authenticated, and then this principal fetches itself from the database to test the administrative daemon, kadmind.

The probe executes the following steps.

1. Probe the ports for kadmind(1M) and krb5kdc(1M) to make sure that they are listening. Run the probe command by using the time-out value that the resource property Probe_timeout specifies. The probe is run every Cheap_probe_interval, which by default is every 30 seconds.

2. Every Thorough_probe_interval (by default 300 seconds) kadmin.local(1M) is used to add a principal. The probe then performs a kinit(1) with the newly created principal. The probe uses the newly created principal to run kadmin(1M) to retrieve its record from the principal database.

3. The result of these probe commands can be either fail or succeed. If Kerberos successfully responds, the probe returns to its infinite loop, waiting for the next probe time.

   If the probe fails, the probe considers this scenario a failure of the Kerberos data service and records the failure in its history. The Kerberos probe considers every failure a complete failure.

4. Based on the success or failure history, a failure can cause a local restart or a data service failover. Refer to Tuning Fault Monitors for Sun Cluster Data Services in Sun Cluster Data Services Planning and Administration Guide for Solaris OS for further details.