Hardware failure causing unscheduled downtime
Our RAID disk array started resynchronization of it's disks resulting in a diminished throughput of the database cluster. This resulted in the eventual cluster crashes. The cluster no longer has limitations, and the resynchronization process has finished.