Jobs & Queues

Background jobs, cron, and durable workflows.

Anything that shouldn't block a request, like sending an email, processing a file, or a scheduled task, belongs in a background job or queue rather than inline in your request handler. These tools range from simple cron scheduling to durable workflow engines that survive process restarts and retries. Match the tool to how critical it is that a job actually completes, not just that it starts.

BullMQ

The go-to if you already run Redis and want a battle-tested job queue library rather than a hosted platform. You own the infra, you keep the control.

↗

Inngest

Write a normal async function, trigger it off an event or cron, and it durably resumes from the last completed step if something crashes mid-run. No separate queue infra to stand up.

↗

Temporal

The heavyweight option once your background work turns into genuinely long-running, multi-step distributed workflows that need to survive crashes for days, not seconds.

↗

Trigger.dev

Similar durable-background-task pitch to Inngest, open source (Apache 2.0). Good if you want the option to self-host later.

↗

Upstash QStash

HTTP-based queue: POST a job to QStash, it retries and delivers to your endpoint later. No worker process to deploy, which is nice on serverless.

↗