How to deploy React + Node.js application to Heroku

In this article, we will explore, Step by step instructions for deploying React + Node.js app to Heroku which we created in my previous article HERE

Let's get started

Clone the repository code from HERE

To run the application, execute following commands

In first terminal

1. yarn install
2. yarn start

In another terminal

1. cd server
2. yarn install
3. yarn start

Once both the servers are started, you can access the application at http://localhost:3000/

Now, let's start with steps to deploy to Heroku.

Heroku requires, the application to have a start script.

Open server/package.json and check if we have a script with name start in package.json file

"start": "nodemon server.js"

Nodemon is good when we are developing on local environment but on Heroku, we don't need nodemon but just node is required.

So remove nodemon entry from dependencies section and change script from

"start": "nodemon server.js"

to

"start": "node server.js"

Now, your package.json will look like this

{
  "name": "server",
  "version": "1.0.0",
  "main": "index.js",
  "license": "MIT",
  "scripts": {
    "start": "node server.js"
  },
  "dependencies": {
    "axios": "0.19.2",
    "cors": "2.8.5",
    "express": "4.17.1"
  }
}

By default, Heroku runs our Node.js application on some randomly available port which is stored in process.env.PORT variable. So we need to make sure, when we start the server we are using that environment variable.

We have already added the code for it in our server.js file

const PORT = process.env.PORT || 5000;

which says take the port value from process.env.PORT environment variable and if the environment variable with name PORT does not exist then use default value of 5000.

Then we are starting our server by accessing the variable PORT

app.listen(PORT, () => {
  console.log(`server started on port ${PORT}`);
});

Now, open main package.json and change

"scripts": {
  "start": "react-scripts start",
  "build": "react-scripts build",
  "test": "react-scripts test",
  "eject": "react-scripts eject"
}

to

"scripts": {
  "start-client": "react-scripts start",
  "build": "react-scripts build && (cd server && yarn install)",
  "test": "react-scripts test",
  "eject": "react-scripts eject",
  "start": "cd server && yarn start"
}

Here, we have done following changes

1. Changed the script name from start to start-client
2. Changed build command
3. Added new start script to start the Node.js server

The reason for specifying yarn install in build command is that, Heroku will run this command while building the app so before running start command, node_modules folder will be available

Now, open src/utils/constants.js and change

export const BASE_API_URL = 'http://localhost:5000';

to

export const BASE_API_URL = '';

Now, stop both the servers which we started initially by running yarn start command.

Now, run the yarn build command which will execute the following command

react-scripts build && (cd server && yarn install)

Now, once the command is executed successfully, run the start script

yarn start

which will execute the following command

cd server && yarn start

Now, you can access the application by visiting http://localhost:5000/
and your Node.js API will also be accessible on the same port 5000 at http://localhost:5000/jobs

The reason for React and Node.js application running on the same port is because when we run the yarn build command, it creates a build folder in our project which contains all the code of our React application and in server/server.js file, we have the following code

const buildPath = path.join(__dirname, '..', 'build');
app.use(express.static(buildPath));

This code will take the contents of build folder and If the 'build' folder contains .html file, then Express server will render the contents of that file when we access http://localhost:5000/.

And our Node.js APIs are already available on Port 5000 so both the applications are running on same port and therefore we don't need to execute two separate commands in separate terminals to start the React App and Node.js application.

So now there is no need of using cors npm module to add as a middleware in server.js

app.use(cors());

This was only required because previously our React application was running on port 3000 and Node.js application was running on Port 5000 and browser does not alllow accessing data from different ports because of Cross-origin resource sharing (CORS) policy. So to allow our Node.js server accept requests from any client, we added the cors package.

But its not required now. So remove the cors package from server/package.json file.

Your updated package.json will look like this now

{
  "name": "server",
  "version": "1.0.0",
  "main": "index.js",
  "license": "MIT",
  "scripts": {
    "start": "node server.js"
  },
  "dependencies": {
    "axios": "0.19.2",
    "express": "4.17.1"
  }
}

Also remove the import for cors in server.js file and its use.

const cors = require('cors');
app.use(cors());

So your final server.js file will look like this

const path = require('path');
const axios = require('axios');
const express = require('express');
const app = express();

const PORT = process.env.PORT || 5000;

const buildPath = path.join(__dirname, '..', 'build');
app.use(express.static(buildPath));

app.get('/jobs', async (req, res) => {
  try {
    let { description = '', full_time, location = '', page = 1 } = req.query;

    description = description ? encodeURIComponent(description) : '';
    location = location ? encodeURIComponent(location) : '';
    full_time = full_time === 'true' ? '&full_time=true' : '';
    if (page) {
      page = parseInt(page);
      page = isNaN(page) ? '' : `&page=${page}`;
    }
    const query = `https://jobs.github.com/positions.json?description=${description}&location=${location}${full_time}${page}`;
    const result = await axios.get(query);
    res.send(result.data);
  } catch (error) {
    res.status(400).send('Error while getting list of jobs.Try again later.');
  }
});

app.listen(PORT, () => {
  console.log(`server started on port ${PORT}`);
});

Now, we're done with the code changes.

Create a new repository in GitHub and let's push the changes to Github

Add all changes to staging area

git add --all .

Commit the changes

git commit -m "Changes for heroku"

Push the changes to github

git push origin master

Now, let's deploy the application to Heroku.

• Login to your Heroku account and navigate to dashboard HERE

• Click on the New button from top right section and select Create new app option

• Provide the name for the App and click on create app button

• Click on Connect to GitHub Link under deployment method section

• Enter name of your GitHub repository and click on Search button

• Click on the Connect button to connect the repository to Heroku

• Select the branch to deploy (It's master in our case) and click on Deploy Branch button

• Now, your deployment will start

• Once it's completed, You will see a success message along with the option to view your deployed App

You can find the complete GitHub source code for this article HERE and live demo of the App deployed on Heroku HERE

Note: For Heroku, If you're using a free account, then your application will sleep after 30 minutes of inactivity and once new request comes, the application will start again so it might take some time to load the application.

Don't forget to subscribe to get my weekly newsletter with amazing tips, tricks and articles directly in your inbox here.