node-local-dns should support graceful shutdown with the ready plugin

[jsws]

We are using node-local-dns with a CiliumLocal Redirect Policy as described here so we need node-local-dns to be able to become Not Ready once it receives a SIGTERM so time is given for the Endpoint to stop receiving traffic and for requests to be directed to kube-dns instead, before the pod is fully terminated. This will allow graceful rollouts of the node-local-dns daemonset.

In #485 it was requested the ready plugin be added to node-local-dns but it was closed on the basis it wouldn't serve any purpose as it would behave the same way as the health plugin. However I believe that's not the case and the ready plugin would allow node-local-dns to gracefully shutdown, since once it receives a SIGTERM the /ready endpoint would start returning 503 which can be used in a Readiness Probe on the pod to determine the pod is shutting down and give time for the Endpoint to be removed when used with the health plugins lameduck setting.
So I am again requesting the ready plugin be added to the plugins list here.

[LeoQuote]

Hi, I’m currently exploring this option( cilium + nodelocal dns), but I see no reason to use image provided by this project instead of using coredns directly, did you try to use coredns directly? Is there any hidden disadvantages?

I tried to use coredns instead and even drafted a PR for this. Hope that would help you coredns/helm#173

[k8s-triage-robot]

The Kubernetes project currently lacks enough contributors to adequately respond to all issues.

This bot triages un-triaged issues according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, lifecycle/rotten is applied
• After 30d of inactivity since lifecycle/rotten was applied, the issue is closed

You can:

• Mark this issue as fresh with /remove-lifecycle stale
• Close this issue with /close
• Offer to help out with Issue Triage

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale