Does draino double check the NodeCondition before proceeding to drain the node

[harighub]

We've condition where draino identified the permanent NodeCondition problem and it was scheduled to drain after 10 mins as configured:

2019-09-30T21:06:40.824Z	DEBUG	kubernetes/eventhandler.go:130	Cordoning	{"node": "10.0.16.126"}
2019-09-30T21:06:40.825Z	INFO	kubernetes/eventhandler.go:139	Cordoned	{"node": "10.0.16.126"}
2019-09-30T21:06:40.825Z	INFO	kubernetes/eventhandler.go:148	Scheduled drain	{"node": "10.0.16.126", "after": "2019-09-30T21:16:40.793Z"}
2019-09-30T21:16:40.793Z	DEBUG	kubernetes/eventhandler.go:152	Draining	{"node": "10.0.16.126"}
2019-09-30T21:16:40.794Z	INFO	kubernetes/eventhandler.go:161	Drained	{"node": "10.0.16.126"}

At the time of draining the system the node doesn't have permanent problem any more and NodeCondition was updated by NPD accordingly. I believe draino should double check whether still has the problem exists.

[negz]

I believe draino should double check whether still has the problem exists.

Definitely a good point - I don't think it does at the moment but I agree it should.