Mark decommissioned nodes as not in service

The VMs that make up a Kubernetes cluster can change over time. We should be able to mark nodes as not in service.

I’ve gone ahead and added an issue to chronograf to help track this feature: Mark decommissioned nodes as not in service