Agent drain: implement agent side. (Remaining: server)

[jkh52]

Agent drain: first half of #566

• Agent sends drain signal to Server.
• Server only logs it.
• Test cases.

Note that if a new agent binary sends the DRAIN message to an old server binary, it is harmless and will be ignored:

apiserver-network-proxy/pkg/server/server.go

Line 994 in 3026615

klog.V(5).InfoS("Ignoring unrecognized packet from backend", "packet", pkt, "agentID", agentID)

[k8s-ci-robot]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: jkh52

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [jkh52]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[cheftako]

Once we begin draining, do we want to timeout and stop if no second signal is sent?

[jkh52]

I think the most cautious (and simple) decision is no, keep an agent in this state fully functioning.

[cheftako]

To be clear this is doing nothing to actually "drain"? Its just a prep step for that later PR?

[jkh52]

To be clear this is doing nothing to actually "drain"? Its just a prep step for that later PR?

Right, this PR should be a functional no-op for the agent, except for sending the draining hint to the server (for later PR: agent selection).

[jkh52]

/assign @cheftako
/assign @tallclair

[cheftako]

/lgtm
Looking forward to the actual drain logic being added. We clearly need to discuss edge cases such as what to do when ALL agents are in the drain state.