OK, problem mostly solved. Appears to be an internal ground failure within the cluster. Will follow up later on with full details, but don't want anyone to waste time with further followups for now!