DevOps / SRE / Platform Engineer / Stage 9 · Operating a cluster
Follow a pod through scheduling to service discovery
"Already know" meets prerequisites. "Not relevant" removes the topic from your total but leaves what builds on it waiting.
Self-check
- In a kind cluster, deploy two replicas behind a Service. Trace a pod from scheduling to running, find it through the Service's DNS name, and explain why a pod with a wrong label receives no traffic.
- Evidence to include: Keep the manifests, the pod events from kubectl describe, the Service endpoints, and the label that was wrong.
Try it without the material first, then compare. Save your answer in your notes below. Report a problem with this self-check
Learning activities
4 activities- Not doneQuick Start (opens another site)
Creating and deleting a local cluster with kind, and loading your own images into it.
4 min - Not donePods (opens another site)
What a pod is and why controllers create them.
13 min - Not doneKubernetes Scheduler (opens another site)
How the scheduler filters and scores nodes before binding a pod.
3 min - Not doneService (opens another site)
Stable endpoints for a changing set of pods, through selectors and DNS.
20 min