tectonic coreos 1.8.9 | pods with shared filesystem ROOK (=ceph ) are stuck in terminating status

[Bert van Houten]

Hi All,

I am not sure but after the update 1.8.9-coreos-1 pods terminated because of a rolling update will not be killed and are stuck in the terminating status forever. 

restarting the kubelet service on the node (s) will correct this.

Sound it familiar for one of you? Do you know and can it be solved?

thx

Bert