IP Address Management

The cluster has been running for a year. A new node joins and no pods can be scheduled on it. Nothing is broken. You ran out of addresses, and the decision that caused it was made at cluster creation.

What this lesson covers

• Pod CIDR allocation per node and how the per-node block size interacts with cluster size
• Address exhaustion as a real production problem rather than a theoretical one
• CIDR sizing decisions made at cluster creation that cannot easily be changed afterwards
• What actually happens when you get it wrong, and what the recovery options cost

What is actually happening

Coming soon.

The Kubernetes model versus the CNI's choice

Coming soon.

How it works in practice

Coming soon.

Common failure modes

Coming soon.

What people get wrong

Coming soon.