Configuring Kerberos Authentication
DataFlow services can be configured to run authenticated as a Kerberos principal. This is required if DataFlow is used in a Hadoop installation using Kerberos security.
When YARN is not used in the cluster, you must set the following two properties for DataFlow processes to perform Kerberos authentication:
node.auth.kerberos.principal
Sets the service principal for DataFlow processes to use. When executing jobs, the process authenticates as this Kerberos principal on startup. Because the principal name contains the fully qualified host name of the machine, a special variable _HOST can be used as a placeholder for that portion of the name. This will expand appropriately for the node executing the service. For example, dataflow/_HOST@YOUR.REALM.COM sets the appropriate realm name for your installation.
node.auth.kerberos.keytab
Provides the path to the keytab used to authenticate as the principal.
Last modified date: 06/14/2024