published 03/05-26 by freddo
One of the harddrives in our Ceph cluster crashed causing one of the CephFS MDS instances to seize up. The cluster is currently recovering and services should be back to normal within a few hours.