Debugging Flaky Network and Timeout Issues
Isolate intermittent connectivity problems.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
ROLE: You are a networking-savvy backend engineer. CONTEXT: Symptoms (timeouts, resets, slow responses) and frequency: [SYMPTOMS]. Architecture between client and server (proxies, load balancers, DNS, service mesh): [ARCHITECTURE]. Logs and metrics available: [OBSERVABILITY]. Recent changes: [RECENT_CHANGES]. STEPS: 1. Build a layered hypothesis list (DNS, TCP connect, TLS, proxy limits, connection pool exhaustion, server overload, retries amplifying load). 2. For each hypothesis name a test: timing breakdown of a request, packet or connection statistics, and comparing paths. 3. Explain how to read timeouts (connect versus read) and idle connection reuse problems. 4. Suggest configuration checks for keep-alive, pool sizes, timeouts, and retries. 5. Propose a reproduction using fault injection and a monitoring dashboard to catch recurrence. OUTPUT FORMAT: Hypothesis table, tests, interpretation guide, configuration checklist, and reproduction plan. QUALITY BAR: Change one variable at a time. Keep evidence from each step. FOLLOW-UP: Ask me for a timing breakdown of a failing request.
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.