Deep Dive Kubernetes Pod Start and Failure Modes

<p><strong>What We Cover in This Episode:</strong></p><p><strong>The Probe Trap,</strong> Why telling an interviewer that a &quot;liveness probe failure removes traffic&quot; is an instant red flag (it actually kills and restarts the container!), and why you should <em>never</em> check external databases in your liveness probes.</p><p><strong>The JWT Myth:</strong> Why saying &quot;JWTs are encrypted&quot; will cost you points. We explain how to articulate that standard JWTs are <em>signed</em>, and how to defend against the notorious alg: none attack.</p><p><strong>Silent Istio YAML Bugs:</strong> We expose the most common structural bug candidates write on the whiteboard: putting fault, retries, and route as separate list items in an Istio VirtualService, which silently fails to route traffic.</p><p><strong>Zero-Trust Security Illusions:</strong> Did you know that Istio&#39;s RequestAuthentication alone does <em>not</em> reject unauthenticated requests? We explain why you absolutely need an AuthorizationPolicy to actually block traffic.</p><p><strong>The Sidecar Evolution:</strong> How to elevate your answer from a mid-level to a Staff-level by explaining the new Kubernetes 1.29 native sidecars (restartPolicy: Always), effectively solving the old startup race conditions</p><p><br></p>