Whenever you're ready, walk me through how you'd approach this.
Explain how you would debug a performance bottleneck in a production environment under heavy load. Describe how you would protect users while investigating, narrow the problem using observability data, distinguish between application, database, infrastructure, and dependency causes, and validate a mitigation. Explain how you would communicate status, decide whether to roll back, scale, throttle, or defer changes, and document follow-up actions after service stability is restored.