I maintain Valkey GLIDE – built a Node queue doing 48k jobs/s
BullMQ replacement hitting 48k jobs/s via 1-RTT ops and Rust native bindings.
A fast in-memory job queue for Node.js
BullMQ API without Redis when you don't need distributed workers.
Node.js developers building small apps and CLIs
BullMQ · pg-boss · node-queue
I have been using BullMq for most of my background job related work but lately I have been working on some smaller scale app where I having a full blown Redis setup with separate worker process would have been overkill.
That is why I built a lightweight, in-memory Node.js background job queue. It comes with a Bull/BullMq-type API, concurrency control, schedulers, job retention and much more. It has zero dependencies and is fully typesafe. Use this when you need a simple, performant background job system but don't need distributed workers and Redis-persistence (best for small apps, CLIs and local tools).
Let me know what you think!
BullMQ replacement hitting 48k jobs/s via 1-RTT ops and Rust native bindings.
Replaces Redis+RQ with just Postgres—fork isolation, transactional enqueueing, zero broker.
Ditches Redis entirely by leveraging Postgres LISTEN/NOTIFY for instant wakeups.
SQLite for local dev without Redis, but Trigger.dev and Inngest already own this space.
Replaces Redis, RabbitMQ, and Kafka with one self-contained Rust binary.
Zero-Redis job queue that handles AI pipeline retries better than Celery.