@quirrel/owl

Owl 🦉

• Getting Started
• What's special about Owl?
• Quirrel 🐿

Owl is a high-performance, redis-backed job queueing library originally built for Quirrel 🐿.

Getting Started

npm install @quirrel/owl

import Owl from "@quirrel/owl"
import Redis from "ioredis"

const owl = new Owl(() => new Redis())

const worker = owl.createWorker(async (job, ackDescriptor) => {
  console.log(`${job.queue}: Received job #${job.id} with payload ${job.payload}.`);
  // ...

  await worker.acknowledger.acknowledge(ackDescriptor);
})

await worker.start();

const producer = owl.createProducer()

await producer.enqueue({
  queue: "email",
  id: "some-random-id",
  payload: "...",
  runAt: new Date(Date.now() + 1000),
  ...
})

While I originally created Owl for use in Quirrel, I decided to publish it as its own project so people can use it for their own purposes. If you want to use Owl in your own project and need some more documentation: Please go ahead and create an issue for it :D

What's special about Owl?

Owl ...

• ... doesn't require you to specify queues upfront
• ... is optimised for short-running jobs
• ... allows for totally custom schedules
• ... is written in TypeScript
• ... has a low-overhead activity stream (based on Redis Pub/Sub)
• ... allows fast queries about currently scheduled jobs
• ... has an persisted mode, but also an in-memory one for quick development

Owl's Architecture

A job consists of a Queue, an ID and a payload.

They are scheduled for later execution by the producer.

Once the time has come for a job to be executed, a worker will request it. This will move it into a list currently processing jobs. Aftere execution is finished, the worker acknowledges it and (in case of repeated jobs) re-enqueues it.

Trivia

Why Owl?

It's well-known that Squirrels 🐿 and Owls 🦉 are good friends. Owls are reliable, mostly down-to-earth and know how to deal with time. Thus, their skillset makes them excellent queue keepers.

Compatibility with Redis Cluster

At the moment, Owl does not aim to be compatible with Redis Cluster. This may change in the future, though.

Contributors ✨

Thanks goes to these wonderful people (emoji key):

Simon Knott 💻 🤔

Antony 💻 🐛

This project follows the all-contributors specification. Contributions of any kind welcome!