Using a daemonset that causes failures during draining as leases are not gracefully released and instead age out as pods are killed after potentially losing network access due to daemonset pods not being terminated. 

As pointed out in https://github.com/openshift/origin/pull/27394#discussion_r964002900 


This should be fixed when moving to a deployment and is also tracked here https://issues.redhat.com/browse/BUILD-495 

100

1. 
2. 
3.