plainjob

A SQLite-backed job queue for better-sqlite3 and bun:sqlite processing 15k jobs/s.

Getting Started

Using bun

bun add plainjob

Create a queue:

import { bun, defineQueue } from "plainjob";
import Database from "bun:sqlite";

const connection = bun(new Database("data.db", { strict: true }));
const queue = defineQueue({ connection });

Make sure strict mode is enabled!

Using node

npm install plainjob better-sqlite3

Create a queue:

import { better, defineQueue } from "plainjob";
import Database from "better-sqlite3";

const connection = better(new Database("data.db"));
const queue = defineQueue({ connection });

Minimal Example

import { bun, defineQueue, defineWorker } from "plainjob";
import Database from "bun:sqlite";

const connection = bun(new Database("data.db", { strict: true }));
const queue = defineQueue({ connection });

// Define a worker
const worker = defineWorker(
  "print",
  async (job) => {
    console.log(`Processing job ${job.id}: ${job.data}`);
  },
  { queue }
);

// Add a job
queue.add("print", "Hello, plainjob!");

// Start the worker
worker.start();

Features

• SQLite-backed: Reliable persistence using bun:sqlite or better-sqlite3
• High performance: Process up to 15,000 jobs per second
• Cron-scheduled jobs: Easily schedule recurring tasks
• Delayed jobs: Run jobs after a specified delay
• Automatic job cleanup: Remove old completed and failed jobs
• Job timeout handling: Re-queue jobs if a worker dies
• Custom logging: Integrate with your preferred logging solution
• Lightweight: No external dependencies beyond better-sqlite3 and a cron-parser

Usage

Creating a Queue

import { bun, defineQueue } from "plainjob";
import Database from "bun:sqlite";

const connection = bun(new Database("data.db", { strict: true }));
const queue = defineQueue({
  connection,
  timeout: 30 * 60 * 1000, // 30 minutes
  removeDoneJobsOlderThan: 7 * 24 * 60 * 60 * 1000, // 7 days
  removeFailedJobsOlderThan: 30 * 24 * 60 * 60 * 1000, // 30 days
});

Adding Jobs

// Enqueue a one-time job
queue.add("send-email", { to: "user@example.com", subject: "Hello" });

// Run a job a after 1 second
queue.add(
  "send-email",
  { to: "user@example.com", subject: "Hello" },
  { delay: 1000 }
);

// Schedule a recurring job
queue.schedule("daily-report", { cron: "0 0 * * *" });

Defining Workers

import { defineWorker } from "plainjob";

const worker = defineWorker(
  "send-email",
  async (job) => {
    const { to, subject } = JSON.parse(job.data);
    await sendEmail(to, subject);
  },
  {
    queue,
    onCompleted: (job) => console.log(`Job ${job.id} completed`),
    onFailed: (job, error) => console.error(`Job ${job.id} failed: ${error}`),
  }
);

worker.start();

Managing Jobs

// Count pending jobs
const pendingCount = queue.countJobs({ status: JobStatus.Pending });

// Get job types
const types = queue.getJobTypes();

// Get scheduled jobs
const scheduledJobs = queue.getScheduledJobs();

Advanced Usage

Graceful Shutdown

To ensure all jobs are processed before shutting down:

import { processAll } from "plainjob";

process.on("SIGTERM", async () => {
  console.log("Shutting down...");
  await worker.stop(); // <-- finishes processing jobs
  queue.close();
  process.exit(0);
});

Multi-process Workers

For high-throughput scenarios, you can spawn multiple worker processes. Here's an example based on bench-worker.ts:

import { fork } from "node:child_process";
import os from "node:os";

const numCPUs = os.cpus().length;
const dbUrl = "queue.db";

for (let i = 0; i < numCPUs; i++) {
  const worker = fork("./worker.ts", [dbUrl]);
  worker.on("exit", (code) => {
    console.log(`Worker ${i} exited with code ${code}`);
  });
}

In worker.ts:

import Database from "better-sqlite3";
import { better, defineQueue, defineWorker, processAll } from "plainjob";

const dbUrl = process.argv[2];
const connection = better(new Database(dbUrl));
const queue = defineQueue({ connection });

const worker = defineWorker(
  "bench",
  async (job) => {
    // Process job
  },
  { queue }
);

void worker.start().catch((error) => {
  console.error(error);
  process.exit(1);
});

This setup allows you to leverage multiple CPU cores for processing jobs in parallel.

For more detailed information on the API and advanced usage, please refer to the source code and tests.