QStash is a robust message queue and task-scheduling service that integrates perfectly with Next.js. This guide will show you how to use QStash in your Next.js projects, including a quickstart and a complete example.
At its core, each QStash message contains two pieces of information:
URL (which endpoint to call)
Request body (e.g. IDs of items you want to process)
The following endpoint could be used to upload an image and then asynchronously queue a processing task to optimize the image in the background.
upload-image/route.ts
import { Client } from "@upstash/qstash"import { NextResponse } from "next/server"const client = new Client({ token: process.env.QSTASH_TOKEN! })export const POST = async (req: Request) => { // Image uploading logic // 👇 Once uploading is done, queue an image processing task const result = await client.publishJSON({ url: "https://your-api-endpoint.com/process-image", body: { imageId: "123" }, }) return NextResponse.json({ message: "Image queued for processing!", qstashMessageId: result.messageId, })}
Note that the URL needs to be publicly available for QStash to call, either as a deployed project or by developing with QStash locally.
Because QStash calls our image processing task, we get automatic retries whenever the API throws an error. These retries make our function very reliable. We also let the user know immediately that their image has been successfully queued.
Now, let’s receive the QStash message in our image processing endpoint:
process-image/route.ts
import { verifySignatureAppRouter } from "@upstash/qstash/nextjs"// 👇 Verify that this messages comes from QStashexport const POST = verifySignatureAppRouter(async (req: Request) => { const body = await req.json() const { imageId } = body as { imageId: string } // Image processing logic, i.e. using sharp return new Response(`Image with id "${imageId}" processed successfully.`)})
.env
# Copy all three from your QStash dashboardQSTASH_TOKEN=QSTASH_CURRENT_SIGNING_KEY=QSTASH_NEXT_SIGNING_KEY=
Just like that, we set up a reliable and asynchronous image processing system in Next.js. The same logic works for email queues, reliable webhook processing, long-running report generations and many more.
We can use QStash to start a background job by calling the publishJSON method.
In this example, we’re using Next.js server actions, but you can also use route handlers.
Since we don’t have our public API endpoint yet, we can use Request Catcher to test the background job.
This will eventually be replaced with our own API endpoint.
src/app/actions.ts
"use server"import { Client } from "@upstash/qstash"const qstashClient = new Client({ // Add your token to a .env file token: process.env.QSTASH_TOKEN!,})export async function startBackgroundJob() { await qstashClient.publishJSON({ url: "https://firstqstashmessage.requestcatcher.com/test", body: { hello: "world", }, })}
Now let’s invoke the startBackgroundJob function when the button is clicked.
To test the background job, click the button and check the Request Catcher for the incoming request.
You can also head over to Upstash Console and go to the
Logs tab where you can see your message activities.
4
Create your own endpoint
Now that we know QStash is working, let’s create our own endpoint to handle a background job. This
is the endpoint that will be invoked by QStash.
This job will be responsible for sending 10 requests, each with a 500ms delay. Since we’re deploying
to Vercel, we have to be cautious of the time limit for serverless functions.
src/app/api/long-task/route.ts
export async function POST(request: Request) { const data = await request.json() for (let i = 0; i < 10; i++) { await fetch("https://firstqstashmessage.requestcatcher.com/test", { method: "POST", body: JSON.stringify(data), headers: { "Content-Type": "application/json" }, }) await new Promise((resolve) => setTimeout(resolve, 500)) } return Response.json({ success: true })}
Now let’s update our startBackgroundJob function to use our new endpoint.
There’s 1 problem: our endpoint is not public. We need to make it public so that QStash can call it.
We have 2 options:
Deploy our application to a platform like Vercel and use the public URL.
Create a local tunnel to test the endpoint locally.
For the purpose, of this tutorial, I’ll deploy the application to Vercel, but
feel free to use a local tunnel if you prefer.
There are many ways to deploy to Vercel, but
I’m going to use the Vercel CLI.
npm install -g vercel
vercel
Once deployed, you can find the public URL in the Vercel dashboard.
Now that we have a public URL, we can update the URL.
src/app/actions.ts
"use server"import { Client } from "@upstash/qstash"const qstashClient = new Client({ token: process.env.QSTASH_TOKEN!,})export async function startBackgroundJob() { await qstashClient.publishJSON({ // Replace with your public URL url: "https://qstash-bg-job.vercel.app/api/long-task", body: { hello: "world", }, })}
And voila! You’ve created a Next.js app that calls a long-running background job using QStash.
5
Error catching and security
QStash is a great way to handle background jobs, but it’s important to remember that it’s a public
API. This means that anyone can call your endpoint. Make sure to add security measures to your endpoint
to ensure that QStash is the sender of the request.
Luckily, our SDK provides a way to verify the sender of the request. Make sure to get your signing keys
from the QStash console and add them to your environment variables. The verifySignatureAppRouter will try to
load QSTASH_CURRENT_SIGNING_KEY and QSTASH_NEXT_SIGNING_KEY from the environment. If one of them is missing,
an error is thrown.
src/app/api/long-task/route.ts
import { verifySignatureAppRouter } from "@upstash/qstash/nextjs"async function handler(request: Request) { const data = await request.json() for (let i = 0; i < 10; i++) { await fetch("https://firstqstashmessage.requestcatcher.com/test", { method: "POST", body: JSON.stringify(data), headers: { "Content-Type": "application/json" }, }) await new Promise((resolve) => setTimeout(resolve, 500)) } return Response.json({ success: true })}export const POST = verifySignatureAppRouter(handler)
Let’s also add error catching to our action and a loading state to our UI.