Whenever you're ready, walk me through how you'd approach this.
Describe your process for debugging a production-level issue.
Explain how you establish impact, collect evidence, narrow the fault domain, and validate both the fix and the absence of regressions. Include how you distinguish symptoms from root cause, communicate during the incident, and decide whether to roll back, mitigate, or deploy a fix.