I am trying to map out costs for a Shared cluster used in our organization. Since Databricks does not store the sessions in all purpose compute or who accessed the cluster, what are some possible options that I can track which notebooks were attached to the cluster during it's uptime?