Getting Started Guides Architecture API Changelog

Choose SDK

Node/Guides/

Reliability

Harden your Node.js task queue for production workloads.

← PreviousTopic Pub/Sub Next →Error Handling

Error HandlingTask failure lifecycle, error inspection, debugging patterns Delivery GuaranteesAt-least-once delivery and exactly-once patterns RetriesAutomatic retries with exponential backoff TimeoutsBound task runtime and reclaim stalled jobs IdempotencyDeduplicate enqueues with unique keys DebouncingCollapse a burst of enqueues into one run on a sliding deadline Flow controlDebounce vs throttle vs coalesce, and what to do with excess load Rate LimitingThrottle task execution with token-bucket limits ConcurrencyCap how many jobs of a task run at once Circuit BreakersProtect downstream services from cascading failures Dead-letter queueCapture and replay exhausted jobs Distributed LockingMutual exclusion across workers with database-backed locks