Reliable n8n Webhook Workflow: HTTP Requests & Data Table

Reliable n8n Webhook Processing with Idempotency, Retries, and Deterministic Responses

This n8n workflow receives POST webhooks, validates and normalizes incoming data, and uses an n8n Data Table to ensure the same eventId is processed idempotently—while calling a downstream HTTP endpoint with bounded retries and backoff. The result: predictable, deterministic webhook responses even during duplicate delivery or downstream failures.

What this workflow does

• Receives incoming POST webhooks via an n8n Webhook endpoint.
• Normalizes and validates payload fields—including eventId, eventType, customerId, amount, and an ISO-8601 timestamp. Invalid payloads return HTTP 400.
• Enforces idempotent processing using the n8n Data Table “Webhook Reliability State” to detect prior eventId entries and return deterministic outcomes for duplicates, in-flight events, and terminal/no-op states.
• On new events, writes a persisted state (e.g., “received”, then marks as “processing”) and triggers a downstream HTTP request to httpbin.org.
• Classifies downstream results into success, retryable failure, or terminal failure, updates the Data Table accordingly, and responds with processed or error details.
• Retries retryable failures with bounded backoff (2 seconds then 4 seconds) up to three attempts, then persists a final failed state and returns an error response.

Use cases

• Reliable webhook handling for SaaS events where duplicates are possible (e.g., event replays).
• Resilient downstream calls when external APIs may temporarily fail.
• Operational confidence for automation engineers needing durable, restart-safe webhook processing.

Technical details

• Uses nodes: webhook, if, set, code, switch, wait.
• Idempotency and durable state tracking with n8n Data Table named “Webhook Reliability State” (keyed by eventId).
• Downstream HTTP calls to httpbin.org with bounded retry/backoff behavior.