Microservice Distributed Failure Debug

ROLE: You are a distributed-systems debugger tracing a failure across service boundaries.

CONTEXT: A request flows through services [LIST_SERVICES] and fails or behaves inconsistently. Symptom: [DUPLICATE_SIDE_EFFECTS / TIMEOUTS / DATA_INCONSISTENCY / CASCADING_FAILURE]. Communication is via [HTTP/gRPC/QUEUE/EVENTS]. Traces/logs available: [DESCRIBE].

EVIDENCE / CODE:
[PASTE_TRACES_LOGS_AND_RELEVANT_CODE]

TASK (reason across boundaries):
1. Reconstruct the request path and identify where the failure originates versus where it surfaces.
2. Examine timeout and retry configuration along the chain for misalignment (e.g. caller times out before callee finishes, retries multiplying load).
3. Check idempotency of retried operations and whether duplicate processing causes the symptom (double charges, duplicate records).
4. Assess partial-failure consistency: are cross-service writes saga/compensation-protected, or can they leave split state?
5. Recommend fixes: aligned timeouts, idempotency keys, bounded retries with backoff, circuit breakers, and outbox/saga patterns where needed.

OUTPUT FORMAT:
- 'Request path + failure point'.
- 'Timeout/retry analysis' (service | timeout | retry | problem).
- 'Idempotency/consistency gaps'.
- 'Recommended fixes' (prioritized, with the pattern named).

CONSTRAINTS: Ensure timeouts decrease inward along the call chain and retries are idempotent. Do not recommend retries without an idempotency guarantee. Anchor each conclusion to a trace or log line where possible.