AI Call Error Log Retry Planner with Capped Exponential Backoff

How it works

1. 1.Set base_delay_ms, max_delay_ms and max_attempts in the config node
2. 2.Emit seven fictional error logs, including one invalid record
3. 3.Validate fields, classify each error and compute the capped exponential delay
4. 4.Route items through If to the retry path or the final-failure path
5. 5.Merge both paths and build one sorted report

Use cases

• Scheduling batch re-runs after overnight rate-limit outages
• Separating permanent errors such as bad API keys before re-running a batch

Requirements

• Logs with item_id, status_code, error_message, attempts and last_attempt_at
• Retry policy values in the Retry Policy Config node
• No API keys needed

Setup and testing in n8n

1. Import the JSON into n8n
2. Run Run Retry Planner manually
3. Check that A-001, A-002 and A-003 are scheduled and that A-003's delay is capped at 30000
4. Check that A-004, A-005, A-006 and A-007 are listed as final failures
5. Change max_attempts or the logs to test edge cases, then replace the sample node with your real log source

Service Value

Ready-made workflow template for automation delivery and service execution.