Hadoop Authentication with FreeIPA for ML Workspaces
CDP uses FreeIPA to provide centralised identity management. FreeIPA combines four identity management capabilities: an LDAP user directory, a Kerberos KDC, a DNS server for shared services, and a shared Certificate Authority. This method of identity management, where your users/groups are maintained in FreeIPA and passwords are authenticated via SSO to Active Directory, provides the infrastructure needed for CDP services, without requiring you to expose your AD over the network.
This procedure is required if you want to run Spark workloads in an ML workspace.