Why v3u

bee-queue, kue, bull they did a great job. Their speed are increased by pipeline and lua scripts. However, there is room for improvement:

• Redis does not natively support job-queue, thus library logic is complex and inefficient
• Pipeline and lua scripts can not completely remove round trip times, for example: library sends blrpoplpush command to server every job.
• It is hard to monitor job's meta data from library: execution time, number of queues, ...
• It is hard to convert the library vto other languages

Benchmark

These basic benchmarks ran 10,000 jobs through each library, Node (v10.5.0), system: MacOS 10.13, ram 8G, processor 2G intel core i5. The number shown are average of 10 runs.

• bee-queue: 3017.1725 ms
• v3u: 2144.2959 ms

Start Cheetah job server

Read document at cheetah

Get started

yarn add v3u

Create a queue:

const { Queue } = require('v3u')
const queue = new Queue('mul', {
  uri: 'cheetah://localhost:1991',
})

Create a worker:

queue.process(({ x, y}, done, progress) => {
  done(null, x * y);
})

Create a producer:

queue.createJob({x: 2, y: 3}, (value) => {
  // job done, value = 6
}, (p) => {
  // job progress, is called multiple times
})