Skip to content

Node.js (Magda)

Estimated Time: 5 minutes
Dependencies: None

This guide will let you experience what it's like to develop with Kelda. We will be working with a real-world, open source, microservices application called Magda. We will deploy a development instance of Magda, fix a bug, and see our local changes take effect immediately on the development server. Let's jump in.


  1. Install the Kelda CLI

    Run the following command to download and install Kelda.

    curl -fsSL '' | sh
  2. Download the Kelda examples repository

    This repository contains the source code and Kelda config for Magda.

    git clone
  3. Configure Kelda to work on the Magda project

    The following command creates a file that's required for the kelda commands to work.

    Most users can just use the defaults by hitting [Enter] for each prompt.

    In the root directory of the examples repository, run:

    ./magda/ --use-demo-cluster

    If the command succeeded, you should see something like:

    Wrote config to /Users/kevin/.kelda.yaml
    Kelda maintains a public cluster for demo purposes. If you've already setup a Kelda cluster, just omit the --use-demo-cluster flag.


  1. Start Kelda

    Start Kelda by running

    kelda dev ./magda/magda-web-server
    The kelda dev process should not be killed during development.

    It's needed to sync local changes, and to keep the tunnels open.

    If you accidentally kill it, you can restart it by re-running the script.

    This command boots Magda, syncs the files in the magda-web-server directory, and creates tunnels for accessing the application.


    Wait for all the services to become Ready. The first boot may take several minutes because the Docker images need to be pulled.

  2. View the Bug

    Open http://localhost:8080 in your browser. You should see a website with a search box.

    Try searching for "Australia" -- a red error message should pop up on the bottom right portion of the screen complaining about a SyntaxError in JSON.

  3. Fix the SyntaxError bug

    Fix this bug by editing your local copy of magda/magda-web-server/src/index.js, and replacing this_value_is_wrong with v0 on line 104.

    You will notice that Kelda will automatically sync your change and restart the affected containers. The Development Status portion of the Kelda terminal application informs you of all of the changes that are occurring in real time.

    Kelda Syncing Info

    If you refresh the page in your browser, and then perform a new search, you should see the SyntaxError go away, and search results begin to appear.


  1. View service logs

    kelda logs allows you to inspect the logs of any running service in your cluster.

    In your terminal, run kelda logs gateway to inspect all of the requests that have been made to your gateway service since you began this guide.

  2. SSH into a service

    kelda ssh allows you to remotely access any service in your cluster.

    In your terminal, run kelda ssh gateway. This will place you into a shell on the gateway service container running inside your Kubernetes cluster.

    Run ps aux to see all of the currently running processes inside the gateway service container. The output should look something like this:

    kelda ssh gateway
    # ps aux
    root           1  0.2  0.7 1220256 57016 ?       Ssl  18:52   0:03 node /usr/src
    root          22  0.0  0.0   4344   776 ?        Ss   19:13   0:00 sh
    root          30  0.0  0.0  17512  2080 ?        R+   19:14   0:00 ps aux

    To exit the SSH session, run exit.


  1. Exit Kelda

    Exit kelda dev by entering Ctrl + c into the terminal window where the Kelda CLI is running.

  2. Delete your development namespace

    Run kelda delete to delete your development namespace so that it stops consuming resources in the remote cluster.

Congratulations! You've now seen how easy it is to use Kelda to make changes to your code base locally and instantly see them reflected in a Kubernetes cluster.

What's Next?

If you're interested in using Kelda for your own application, you should deploy Kelda on your own infrastructure next.