Running a Kubernetes app using remote development

This page shows you how to get started quickly with Cloud Code using a remote development environment in Cloud Shell.

In this quickstart, you'll skip setup and clone a project into your remote development workspace with the click of a button, create a cluster, run a Kubernetes app on this cluster, debug your running code, view logs from your live application, and connect a terminal to your running container.

If you're running this quickstart on Apple M-series silicon, you're prompted to build your application using Cloud Build, which can incur charges. For more information, see Cloud Code support for Apple M-series silicon.

Before you begin

To set up supporting resources and access Cloud Shell, the remote development environment on Google Cloud that you use in this quickstart, inside VS Code, complete the following steps:

  1. Sign in to your Google Cloud account. If you're new to Google Cloud, create an account to evaluate how our products perform in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.
  2. In the Google Cloud Console, on the project selector page, select or create a Google Cloud project.

    Go to project selector

  3. Make sure that billing is enabled for your Cloud project. Learn how to confirm that billing is enabled for your project.

  4. Install Visual Studio Code on your machine, if you haven't already.
  5. Install the Remote - SSH Visual Studio Code extension.
  6. Install Git. Git is required for copying samples to your machine.

Launching Cloud Code

When you open Visual Studio Code connected to a remote development environment in Cloud Shell, the environment has all the tools you need to develop Kubernetes applications.

  1. To start work in your remote development workplace, choose your preferred language for the cloned project and then click the Open with Cloud Code button:

    VS Code launches and clones a project into your remote development workspace.

  2. If you don't already have the prerequisites set up, you're prompted to install them.

Creating a GKE cluster

To create a Standard Google Kubernetes Engine (GKE) cluster, follow these steps:

  1. To open the Kubernetes Explorer, click Kubernetes Explorer icon Cloud Code - Kubernetes.
  2. In the header of the Kubernetes Explorer pane, click + Add a Cluster to the KubeConfig.
  3. Choose Google Kubernetes Engine and then click + Create a new GKE Cluster.
  4. Choose Standard as the cluster type.
  5. Click Open to permit Cloud Code to open the Cloud Console.
  6. In Cloud Console, use the project you created, set the zone to us-central1-a, and set the cluster name to my-first-cluster.
  7. Click Create. Cluster creation takes a couple minutes.
  8. After the cluster is created, in Kubernetes Explorer, click Refresh Kubernetes cluster view Refresh.
  9. After the name of your new cluster appears in the list, click the cluster name. Your new cluster is added to the configuration and configured to be the active context.

Running and viewing your app

Now that you're all set up, you can run your application and view it live. Cloud Code watches your filesystem for changes so that you can edit and rerun your app in near real time.

To run your application, follow these steps:

  1. Use the Cloud Code status bar and select Run on Kubernetes.
  2. Confirm whether to use the current cluster context or switch to a different one.
  3. If prompted, choose an image registry to push the images to.

    An output window appears where you can track the progress of your running application. You'll also see a live stream of the logs from the running pods in the terminal output.

  4. After your application is running on Kubernetes, the output window displays an IP address. To use this linked IP address to access your application, press Ctrl/Cmd + click.

Troubleshooting tips

If you're using a pre-existing cluster, to set your cluster as active and get cluster credentials using the Kubernetes Explorer, follow these steps:

  1. To open the Kubernetes Explorer, click the Cloud Code - Kubernetes Kubernetes Explorer icon icon in the VS Code Activity bar on the left.
  2. In the Kubernetes Explorer, right-click your cluster name and then click Set as Active Cluster.

Debugging your app

To debug your application, follow these steps:

  1. Use the Cloud Code status bar and then select Debug on Kubernetes.
  2. If prompted, confirm whether to use the current cluster context or switch to a preferred one.

    Cloud Code uses the cloudcode.kubernetes configurations in your .vscode/launch.json file to run your application and attach a debugger session to it.

    Cloud Code builds your containers, pushes them to the registry, applies Kubernetes configurations to the cluster, and returns the IP address that you can use to browse your live application.

  3. Before your debugger session is attached, you're prompted to confirm or enter the directory in the remote container where the program to debug is found or press ESC to skip debugging the container.

  4. To add a breakpoint to line #9, open src/app.js and then click in the editor margin.

    Red filled circles signify active breakpoints, while gray hollow circles signify disabled breakpoints. For finer breakpoint control, you can use the Breakpoints section in VS Code's Debug view.

    Breakpoints section in the left hand panel of Debug View that allows adding, removing, and disabling breakpoints

    When you send a new request to your application, the debugger pauses at the first active breakpoint.

    In the following sample, under Local in the Variables section, note that the value of res._contentLength; for Hello, world! is 13.

    App paused at breakpoint and variables and call stack sections populated with values in scope

  5. Edit the string being sent in line #8 to Hello, goodbye! and then restart the Debug on Kubernetes action.

    After the app is rebuilt and redeployed, note the updated value of res._contentLength.

    App paused at breakpoint with updated values

Viewing logs

In addition to seeing a live stream of the logs from running pods in the terminal output from your running application, you can also view logs with the Log Viewer that comes with Cloud Code.

To view logs, follow these steps:

  1. Launch the Log Viewer by typing Cloud Code: View Logs using the Command Palette (accessible with Cmd/Ctrl+Shift+P or from the Manage menu Manage icon).
  2. Search for the running app, in this case nodejs-hello-world, to view logs from using the deployment field in the Log Viewer search box.

    Viewing logs by setting the deployment field within the Log Viewer search box to 'nodejs-hello-world'.

Bonus: Opening a terminal in your container

To open a terminal in your container, follow these instructions:

  1. To open the Kubernetes Explorer, click Kubernetes Explorer icon Cloud Code - Kubernetes.
  2. Select the cluster and, from the underlying Pods section, the pod you'd like to connect to.
  3. Under your specified pod, expand your Containers section.
  4. Right-click the container in which you'd like to open a terminal and then click Get Terminal.

    This launches a terminal; you now have access to a shell inside the running container!

    Opening a terminal for the specified pod

Cleaning up

After you terminate your application, all Kubernetes resources deployed during the run are deleted automatically.

To avoid incurring charges to your account for other resources used in this quickstart, be sure to delete the cluster and project you created.

If you're using Google Cloud and would like to delete just your cluster, you can do so by following these steps:

  1. In the Kubernetes Explorer, hover over your cluster name and then click Open in Cloud Console icon Open in Cloud Console.
  2. Click Delete and then click Delete.

To delete your project (and associated resources, including any clusters):

  1. Go to the Projects page in the Cloud Console:

    Go to the Projects page

  2. Select the project that you created for this quickstart and then click Delete.

  3. Type the project ID to confirm and then click Shut down.

    This shuts down the project and schedules it for deletion.

What's next