The process of deploying an Angular 6+ application to Azure is pretty Byzantine, but here are the basic steps to get it going.

I really like Microsoft Azure, but the process of deploying an Angular 6+ app is complex, and filled with some very big hidden gotchas. As of this writing, you cannot just spin up a web app instance, push Angular code to it from a repository and have it run; you need to take several steps first.

Here are the essential steps. Note that another and in many cases better approach is to use containers, such as Docker, but I’m not going to cover that here. In this post, I’ll walk through running your Angular app directly on an Azure web app instance.

Before you begin

Make sure your production build is working on your local machine:

This command should create a dist folder right off the root of your Angular project. If it doesn’t build on your machine, it’s not going to build once it’s deployed to Azure. Normally in development you’re using:

but this only runs Angular on the development server, which is not suitable for production environments.

Jot down a couple things from your local build machine. Let’s find out which version of Node you are running (and npm version), because you’ll want to replicate it on Azure. On your local machine, open up a command shell, and type:

In my case, I get: v10.6.0 at this writing. So I’ll be remembering “10.6.0” as a value when I tell Azure what version of Node I want it to use in the “Application Settings” step below.

Make these modifications to your local Angular app:

  • Azure’s going to need to know how to build your app. To do this, open up package.json and copy a few “devDependencies” into the main root “dependencies” section. Here are the ones that were important to my project’s successful on-Azure build:

    While still in package.json, look at the “scripts” section toward the top of the file. Make you have an entry for “build” : “ng build –prod”. You’ll see in the deploy.cmd file below that we rely upon that command to do the actual build on Azure. You do not commit the dist folder to GitHub; rather, you commit the code source, and Azure first copies over the GitHub code, then it runs an ng build –prod, and it copies the results of the dist folder to the Azure production directory that you define in Application Settings.

    Note too that the “start” command was removed, which seems to confuse Azure; you don’t want Azure trying to “ng serve” your application, because that would try to invoke the development server, which is not suitable for production.

Let’s Create an Azure App

We need a Web App instance to work with. We will then configure the Application Settings to tell it which version of Node to run, what kind of deployment script to run after it sucks in all those files from GitHub when you publish a commit, and we will tell Azure where to look for the app directory.

  1. Using the Azure Portal, create a web app.
  2. Once that’s deployed, go to the Azure Application Settings for that web app, and update two key variables:
    1. App setting name: “WEBSITE_NODE_DEFAULT_VERSION” Set this value to indicate what version of Node Azure should spin up for you. I like to match my local dev machine exactly (nope, folks, I’m not yet using Docker for small deployments.) You can find this value on your local machine by opening up a command prompt and typing “node -v”.For me, at this writing, it’s 10.6.0, so that’s what I set the value of this variable to in Azure.
    2. You also need to tell Azure where to find the code to run, and it’s not going to save in the default directory — it’s going to save in the default directory appended with your project name. At the very bottom of the Application Settings area, you’ll see a section called “Virtual applications and directories”. You need to tell Azure where to go for the main virtual path of your site. It will generally be site\wwwroot\{your-angular-project-name}   That is, that’s what it will be if you follow the custom DEPLOY.CMD command below.
  3. Next, you’ll need to teach Azure’s web service how to build and deploy your app, once it pulls in all that code from GitHub.

    To do this, you create two files — one called “.deployment” and the other called “deploy.cmd.”

    In root level of your Angular project, you’ll need to create a custom deploy.cmd file which builds the “dist” folder and uses that as the main production code.

    To do this, we want to use a custom deploy.cmd, not just the standard Azure one. It does a production build, and then uses the results of the standard Angular dist folder as the main website code.

    Here’s what the files look like. The first filename .deployment (note the period before the word), and it belongs in the root of your Angular project. This simply looks like this:

    The second filename is the actual deployment script, called deploy.cmd, and it looks like this below. Note in particular line 61, which copies from the dist folder after completing that production build.

    Note that this file below not the plain vanilla deploy.cmd — it’s been slightly modified.

    Key Gotcha: both .deployment and deploy.cmd need to go in the ROOT of your GitHub Repository. (e.g., if you’ve got an Angular app project that’s a subfolder of your GitHub repo, still make sure you put those files in the ROOT of your GitHub repo, not the root of the Angular sub project.)

    The last steps are fairly straightforward:

  4. Get your Angular App hosted in a GitHub repository. There are several articles about this, here’s one.
  5. Tell Azure to deploy the app from GitHub (article.) Essentially, you go to the the Azure Portal, choose the web app, go to “Deployment” and set up “Deploy from GitHub.” It’s pretty straightforward if you already have your code in GitHub.

That’s it! Once you do those steps, you should then be able to get code working locally, commit it to GitHub, wait for a few minutes as Azure fetches the code from GitHub and builds it, and then serves it up on request to users visiting your website. Note again that it’s doing a production build, which is a higher level of code quality checking than the debug builds, so if your build suddenly starts failing on Azure, it’s probably not Azure’s fault, it’s probably because your production build doesn’t build locally. Best to do an ng build –prod before deploying to GitHub.

Of course, this is not a “best practice” for production builds — you should be deploying to one or more staging slots, running your test suites on the staging site, and flipping the slots into production, among other things, but that’s not in the scope of this post.


If you’re hosted on IIS, you’ll also want this web.config file shipped up to your Azure host so that Angular can take over the routing.

Troubleshooting Tips

  1. In the “Deployment” blade for your app on Azure, you should see it fetching code from GitHub when you do a checkin to the configured branch. If it doesn’t you might try detaching the connection and re-attaching the GitHub deployment.
  2. If you see it building and then giving a “Failed” message, check the Logs link and scroll through the output. If it says it cannot find any files in the d:\…\dist directory, it means the build did not succeed on Azure. The log file, as cluttered as it sometimes can be, should tell you about component (or components) that are missing. Add those to the “dependencies” section of package.json and check in a new build. Sometimes the first deployment is a bit of an iteration dance, because it will usually fail after the first one and not report all the missing libraries. Think of Azure as basically starting from scratch in the build process — it starts with a clean working directory each time it pulls builds from GitHub, then runs an ng build –prod to build the dist folder.

Steve's an entrepreneur and software leader. Most recently, he founded, a way to create celebration cards easily. He also founded, helping people discover and share productive ways they can respond in times of crisis. Steve's worked on consumer apps, online travel, games, relational databases, management consulting and telecom. He launched Alignvote in 2019, which helped Seattle voters find their best-match political candidates. Steve founded BigOven, the first recipe app for iPhone, now with more than 15 million downloads, which was purchased in 2018. Steve served as Chairman of Escapia Inc., the leading SaaS solution for the US vacation rental industry, sold to Homeaway, now part of Expedia. In 1997, Steve was cofounder, President, CEO and Chairman of VacationSpot, a pioneer in the online reservation of vacation rentals, bought by Expedia in January 2000. At Expedia, Steve was Vice President of Vacation Packages, leading the vacation package and destination services teams, helping to create two patents on the first-ever dynamic vacation packaging system on the Internet, which now represents billions in annual transactions for Expedia. He has keynoted on several occasions at the Vacation Rental Managers Association (VRMA), and taught a graduate level course on the strategic management of innovation at the University of Washington Foster Business School in Seattle, Washington. Steve worked for Microsoft from 1991 to 1997 in a variety of senior marketing and executive positions, and led the creation of the internet games group, helping develop several products and patents related to online multiplayer gaming. He helped launch Microsoft Access and was involved in the acquisition of Fox Software by Microsoft in 1993. He's worked for IBM, Booz-Allen Hamilton and Bell Communications Research. He holds an MS in Computer Science from Stanford University in Symbolic and Heuristic Computation (AI), an MBA from Harvard Business School, where he was named a George F. Baker Scholar (awarded to top 5% of graduating class), and a dual BS in Applied Mathematics / Computer Science and Industrial Management from Carnegie Mellon University (CMU) with University Honors. Steve volunteers when time allows with Habitat for Humanity, University District Food Bank, YMCA Seattle, Technology Access Foundation (TAF) and other organizations in Seattle.

Write A Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.