What is a common mistake with Kubernetes observability and troubleshooting?

Looking only at Pod logs while ignoring events, resource pressure, rollout history, networking, or dependency metrics can hide the real cause.