This guide shows you how to deploy Spartacus apps on Layer0.

This framework has a connector developed for Layer0. See Connectors for more information.

Layer0 only supports Node.js version 14.x

If you do not have Node.js installed on your system, download and install it from the official Node.js v14.x downloads page. Select the download that matches your operating system and run the installer. Note that the installer for Node.js will also install npm.

Note that while you can use any version of Node.js >= 14 locally, your app will run in Node 14 when deployed to the Layer0 cloud. Therefore, we highly suggest using Node 14 for all development.

If you don't already have a Spartacus application, you can create one using:

Spartacus 2.x only supports Angular version 9.x Spartacus 3.x only supports Angular version 10.x

npm install -g @angular/cli@9
ng new my-layer0-spartacus-app

You should now have a working starter app. Run ng serve to see the application running on localhost:4200.

To deploy your Spartacus application on Layer0 it needs to support server-side rendering (SSR). To add SSR support, run:

ng add @spartacus/schematics --ssr

Read more about server-side rendering in Spartacus here.

The previous command created:

  • A server-side application module (app.server.module.ts)
  • A bootstrapper for the server app (main.server.ts)
  • server.ts which exports an Express app
  • TypeScript configuration for the server (tsconfig.server.json)

You can now run npm run build:ssr && npm run serve:ssr to access your server-side rendered app at localhost:4000.

To prepare your Spartacus application for deployment on Layer0:

npm install -g @layer0/cli
  1. Run the following in the root folder of your project. This will configure your project for Layer0.
layer0 init

This will automatically add all of the required dependencies and files to your project. These include:

  • The @layer0/core package
  • The @layer0/angular package
  • The @layer0/cli package
  • The @layer0/spartacus package
  • The @layer0/prefetch package
  • layer0.config.js- Contains various configuration options for Layer0.
  • routes.js - A default routes file that sends all requests to the Angular Universal server. Update this file to add caching or proxy some URLs to a different origin.
  • The sw folder - Contains the files needed to build the service worker that that provides static asset and API prefetching.

For an app called my-layer0-spartacus-app the Layer0 config file created by layer0 init will look like so:

// This file was automatically added by layer0 deploy.
// You should commit this file to source control.

module.exports = {
  backends: {
    commerce: {
      domainOrIp: '',
      hostHeader: '',

If you have several projects and the defaultProject as specified in angular.json is not the project with the SSR build, specify the correct project with the ANGULAR_PROJECT environment variable. For example: ANGULAR_PROJECT=my-ssr-project layer0 build.

The baseUrl should be updated to use the remote URL when window is not defined (i.e., for SSR), and the current host when window is defined. For example:

baseUrl: typeof window !== 'undefined'
  ? `${window.location.protocol}//${}`
  : ''

This value is defined in the backend property of the options parameter to B2cStorefrontModule.withConfig({}) in the app.module.ts file, but is best set using environment variables in the environment.ts and files.

Prefetching for a Spartacus app can be enabled by listening to upstream requests made when server-side rendering a specific page. @layer0/prefetch library will pick up on the upstream requests made by reading the x-0-upstream-requests response header. An example scenario:

  1. User A lands on /product/1.
  2. /product/1 has not been cached in the edge and thus will be server-side rendered.
  3. The rendering server has been modified to track upstream requests by patching https.request.
  4. The rendering server sets x-0-upstream-requests to, for example: /rest/v2/1;/rest/v2/2;
  5. The HTML response for /product/1 is now cached and for future requests served from the edge along with the x-0-upstream-requests response header.
  6. User B lands on a page that has a link to /product/1. /product/:path* has been configured with true. Because of this configuration, @layer0/prefetch will know to make a prefetch HEAD request for /product/1, and only if product/1 can be served from the edge will it prefetch all requests specified in x-0-upstream-requests response header.
  7. When User B click the link to /product/1, the navigation will be faster since the requests needed to render the new page will be in service worker cache.

Example implementation of upstream request tracking changes required in your server.ts file:

import 'zone.js/dist/zone-node'
import * as express from 'express'
import { join } from 'path'

+ // layer0
+ import * as http from 'http'
+ import * as https from 'https'
+ import createRenderCallback from '@layer0/spartacus/server/createRenderCallback'
+ import installLayer0Middleware from '@layer0/spartacus/server/installLayer0Middleware'

// Express server
const server = express()

+ installLayer0Middleware({ server, http, https });

const PORT = process.env.PORT || 4200
const DIST_FOLDER = join(process.cwd(), 'dist/<your-project-name>')

// * NOTE :: leave this as require() since this file is built Dynamically from webpack
const {
} = require('./dist/<your-project-name>-server/main')

    bootstrap: AppServerModuleNgFactory,
    providers: [provideModuleMap(LAZY_MODULE_MAP)],

server.set('view engine', 'html')
server.set('views', DIST_FOLDER)

  express.static(DIST_FOLDER, {
    maxAge: '1y',

// All regular routes use the Universal engine
server.get('*', (req, res) => {
    { req },
+   createRenderCallback(res),

export default server

Some CDNs, such as Akamai, impose low limits on the size of response headers. Prefetching works by listing all of the upstream API URLs fetched during SSR in a x-0-upstream-requests response header. If your application makes many upstream requests for each page during SSR, this header can be quite long and exceed the maximum length allowed by your CDN. To mitigate this, using the maxHeaderLength option when calling createRenderCallback:

createRenderCallback(res, { maxHeaderLength: 500 })

The build command places the built service-worker.js under dist so @layer0/angular will know to static serve the file.

Installing the service worker and any further prefetching will be handled by @layer0/prefetch by invoking the install function imported from @layer0/prefetch/window/install.

Example implementation in app.component.ts:

import { Component, OnInit, Inject } from '@angular/core'
import { isPlatformBrowser } from '@angular/common'
import { PLATFORM_ID } from '@angular/core'
+ import install from '@layer0/prefetch/window/install'

  selector: 'app-root',
  templateUrl: './app.component.html',
  styleUrls: ['./app.component.scss'],
export class AppComponent implements OnInit {
  isBrowser: boolean
  title = '<your-project-name>'

  constructor(@Inject(PLATFORM_ID) platformId: Object) {
    this.isBrowser = isPlatformBrowser(platformId)

  ngOnInit() {
+   setTimeout(() => {
+     if (this.isBrowser) {
+       install()
+     }
+   })

To avoid Spartacus installing ngsw-worker, set production: false in as a temporary workaround:

pwa: {
- enabled: environment.production
+ enabled: false

You may also need to disable it in your app.module.ts file:

-   enabled: environment.production,
+   enabled: false

Add "skipLibCheck": true, to tsconfig.json to avoid type errors from workbox library during build.

The default routes.js file created by layer0 init sends all requests to Angular server via a fallback route.

// This file was automatically added by layer0 deploy.
// You should commit this file to source control.

import { Router } from '@layer0/core/router'
import { angularRoutes } from '@layer0/angular'

export default new Router()
  // other routes removed

The default router also includes common cache configurations for most Spartacus apps:

  return new Router()
    .match('/rest/v2/:path*', ({ cache, proxy }) => {
        browser: {
          maxAgeSeconds: PAGE_TTL,
          serviceWorkerSeconds: PAGE_TTL,
        edge: {
          maxAgeSeconds: PAGE_TTL,
          staleWhileRevalidateSeconds: PAGE_TTL,
      return proxy('commerce')
    .match('/medias/:path*', ({ cache, proxy }) => {
        browser: {
          maxAgeSeconds: FAR_FUTURE_TTL,
        edge: {
          maxAgeSeconds: FAR_FUTURE_TTL,
          staleWhileRevalidateSeconds: 60 * 60 * 24,
      return proxy('commerce')

These routes are set up to cache the default API endpoints from SAP Commerce Cloud, but should be configured to suit your application as needed.

Finally, to configure prefetching for your pages, configure the routes that use SSR using the prefetchUpstreamRequests: true flag for the cache function:

const CACHE_SSR_PAGE = {
  prefetchUpstreamRequests: true,
  edge: {
    maxAgeSeconds: PAGE_TTL * 365,
    staleWhileRevalidateSeconds: PAGE_TTL * 365,
    forcePrivateCaching: true,

return new Router()
  .get('/base-site-path/:path*', ({ cache }) => {

To test your app locally, run:

layer0 run

You can do a production build of your app and test it locally using:

layer0 build && layer0 run --production

Setting --production runs your app exactly as it will be uploaded to the Layer0 cloud using serverless-offline.

Deploying requires an account on Layer0. Sign up here for free. Once you have an account, you can deploy to Layer0 by running the following in the root folder of your project:

layer0 deploy