Grading Checklist

• Identifies that kubectl drain refuses to evict DaemonSet pods by default because they would be immediately rescheduled to the same node.
• Recommends using --ignore-daemonsets flag to skip DaemonSet-managed pods during drain.
• Explains that DaemonSet pods are not evicted because the DaemonSet controller would recreate them on the same node.
• Notes that DaemonSet pods will be removed when the node is actually deleted from the cluster.
• Distinguishes between DaemonSet pods and regular pods in the drain behavior.
• Mentions checking for local storage (emptyDir, hostPath) on DaemonSet pods that might need --delete-emptydir-data.
• Considers whether Fluentd buffers need to be flushed before node removal.
• Provides the correct full drain command with appropriate flags.
• Notes that cordoning the node first prevents new pods from being scheduled.
• Warns against using --force unnecessarily when --ignore-daemonsets is sufficient.
• Mentions verifying all non-DaemonSet pods have been successfully rescheduled after drain.