Our cluster engineers have identified the problem and are bringing affected slices back online.  We expect full resolution shortly, and will update when we confirm all slices on node 17 are back online.