It has never been easier to deploy a next.js application to AWS Lambda!
next-serverless
handles everything for you and deploys your next.js application to AWS Lambda with minimal or even no configuration using the serverless framework.
npm install --save next-serverless
During install, a default serverless.yml
gets created if you don't already have one (if not, copy it from here).
You might want to update it to your needs.
If you are not using a custom server, replace the next
binaries in your package.json
with next-serverless
:
{
"scripts": {
- "dev": "next",
- "build": "next build",
- "start": "next start"
+ "dev": "next-serverless",
+ "build": "next-serverless build",
+ "start": "next-serverless start"
}
}
Or if you are using a custom server, please read using a custom server below.
You are now already good to go and your application is ready for the first deployment (serverless commands) to AWS Lambda.
However, if you are using a randomly generated ApiGateway hostname (e.g. nnptap4nw2.execute-api.eu-west-1.amazonaws.com/prod
) and you are not using a custom (sub-) domain, please read using ApiGateway hostnames as some small changes are necessary.
In this section, we cover a few required changes if you want to use the randomly generated ApiGateway hostnames (e.g. nnptap4nw2.execute-api.eu-west-1.amazonaws.com/prod
).
If you are always using a dedicated (sub-) domain (e.g. myproduct.com
or my.product.com
), you don't have to perform these changes.
Please note that it is not possible to have a mix: a custom (sub-) domain and a stage prefix (/prod
) which you have with ApiGateway hostnames.
This is a limitation by next.js as it, unfortunately, doesn't support path prefixes.
If you are using the randomly generated ApiGateway hostnames, you may have noticed that they have the stage as a path prefix (e.g. nnptap4nw2.execute-api.eu-west-1.amazonaws.com/prod
).
next-serverless
provides a link component which automatically detects those links and automatically adds the stage prefix if required.
This link component has exactly the same signature as the default next.js link component, so you can simply change the imports:
-import Link from 'next/link';
+import Link from 'next-serverless/link';
export default () => (
<Link href="/about">
<a>I will automatically have the stage prefix if needed</a>
</Link>
);
When you are including assets in your app (e.g. /_next/static/style.css
for CSS or an image /_next/static/my-image.png
), you would also get a 404 because it doesn't include the stage prefix.
next-serverless
provides a few helper functions which you can use in these cases:
You can pass in any path you want which then gets prefixes if needed. It automatically takes care of full URLs (https://example.com/styles.css
) and doesn't prefix these.
It returns either the path prefix ('/prod'
) or an empty string (''
) if none is needed for the current request.
If you are using this function instead of prefixPath
, you have to take care of full URLs by yourself.
// ./pages/_document.js
import Document, { Head, Main, NextScript } from 'next/document';
import { prefixPath, getPathPrefix } from 'next-serverless';
export default class MyDocument extends Document {
render() {
return (
<html>
<Head>
<link rel="stylesheet" href={prefixPath('/_next/static/style.css')} />
{/* or */}
<link rel="stylesheet" href={getPathPrefix() + '/_next/static/style.css'} />
</Head>
<body>
<img src={prefixPath(require('./my-image.png'))} />
{/* or */}
<img src={getPathPrefix() + require('./my-image.png')} />
<Main />
<NextScript />
</body>
</html>
);
}
}
If you end up using these helper functions a lot, you may want to create a wrapper component.
For example, create a new Image
component which automatically prefixes the src
attribute.
There are a few changes needed to perform if you are using a custom server.
First of all, your server file needs to export a handler function which AWS Lambda will then call.
next-serverless/handler
provides you a wrapper function for that so you don't have to handle much on your own.
Please note that you should not call app.prepare()
by yourself anymore, next-serverless
will automatically do it at the right time.
You can just use the callback or handler functions provided.
Parameters:
This needs to be an instance of the next.js server/app which you create with next({ /* options */ })
.
The handler
function you provide here will be called for every request.
You can just pass in the default next.js request handler or use your own function here to fully customize the requests and responses.
You can use the normal req
and res
objects of an HTTP Node or express server, we automatically map the Lambda event object to them.
When you run the app locally and not in AWS Lambda, you still want to start your own development server.
You should do that within this callback function.
app.prepare()
already got called before this function gets executed so you don't have to do this again.
const { createServer } = require('http');
const { parse } = require('url');
const next = require('next');
const nextServerless = require('next-serverless/handler');
const dev = process.env.NODE_ENV !== 'production';
const app = next({ dev });
const handle = app.getRequestHandler();
const requestHandler = (req, res) => {
const parsedUrl = parse(req.url, true);
const { pathname } = parsedUrl;
// implement a custom behavior for /hello
if (pathname === '/hello') {
res.end('hello world');
// let next.js handle all other requests
} else {
handle(req, res, parsedUrl)
}
};
// export the nextServerless function
module.exports.handler = nextServerless(app, requestHandler, () => {
// create a normal http node server for local usage
createServer(requestHandler).listen(3000, (err) => {
if (err) throw err;
console.log('> Ready on http://localhost:3000');
});
});
There are also a few more examples (custom routing, using next-routes
or express
) in the examples directory.
If you need assistance to get your custom server running, please feel free to create an issue. More examples are also welcome with a PR.
The last step you have to do when using a custom server is updating your serverless.yml
file.
In functions.app.handler
, we specify the file and exported function which AWS Lambda should execute for every request.
By default, we use the provided server by next-serverless
but since you are using a custom server, you also want to change that.
Simply change this value to the location and exported function of your custom server using the standard Lambda syntax.
For example, if you have a server.js
in your root folder, simply use server.handler
.
next-serverless
is fully compatible with the serverless framework so you can use the same commands here.
You can either install serverless
locally, globally (npm install --global serverless
) or use npx
which comes with npm 5.2+ and higher.
A few example commands:
# build the application before deploying it
$ npm run build
# deploy to the default stage
$ npx serverless deploy
# deploy to a specific stage
$ npx serverless deploy --stage prod
# get the server logs
$ npx logs --stage prod -f app
# watch the server logs
$ npx logs --stage prod -f app -t
MIT © Cyril Wanner