Execution Engine Queue Mode And Scaling

Operator runs n8n in a single Docker container under load. Workflow latency spikes because the main process is both serving the UI and running long workflows.

Switch to EXECUTIONS_MODE=queue with a Redis URL and one or more `n8n worker` processes. Main process handles UI + trigger registration; workers consume jobs. Pin worker concurrency via N8N_CONCURRENCY_PRODUCTION_LIMIT per worker. Do not leave both modes mixed across instances.

A workflow includes a Code node with an accidental infinite loop. The worker hangs indefinitely.

Set EXECUTIONS_TIMEOUT (seconds) at the instance level and EXECUTIONS_TIMEOUT_MAX as the operator-facing ceiling. Set per-workflow `executionTimeout` for long legitimate jobs. Combine with Code-node sandbox to abort runaway scripts.

A queue-mode worker OOMs in the middle of a long HTTP node. Bull marks the job stalled.

n8n's behavior on stalled jobs depends on Bull's stalled-job recovery + the operator's QUEUE_RECOVERY_INTERVAL [REQUIRES-VERIFICATION]. Verify whether the workflow is retried from scratch, marked failed, or resumed at the node. Design workflows to be idempotent end-to-end (esp. write nodes) so retr…