No incidents reported
Full cluster healthy again.
Server auto-management of memory resolved the issue.