Problem Statement: We are currently utilizing customer-managed keys for Databricks compute encryption at the workspace level. As part of our key rotation strategy, we find ourselves needing to bring down the entire compute/clusters to update storage encryption keys. However, we encounter errors when attempting to update storage encryption keys without shutting down the compute.
Our workspace is shared by multiple application teams, each with automated jobs triggering compute/clusters to start. The process of stopping all workflows/jobs manually is time-consuming. Is there a way to temporarily pass access at workspace level and allow only Databricks admins to facilitate this key rotation process
Any guidance or best practices on handling key rotations in a shared workspace environment would be greatly appreciated.
Thank you.