Kured should verify if node is still drained when delaying reboot

[sebastiangaiser]

Hey,

we had a problem with two controllers draining/uncordon nodes. To make the reboot more graceful, we configured rebootDelay: 120s. During these 2 mins the other controller uncordened the node. Kured rebooted the not drained node afterwards which is not optimal :D So Kured should verify after the delay and right before the reboot is issued, if the node is still drained or not.