Data Management with Nuvla Part 3

11 August 2020 | by Konstantin Skaburskas

edge cloud

Implementing a successful edge computing strategy requires close cooperation between software developers and IT operations. Adopting DevOps principles is therefore the natural way to go. In this blog, we explore how the Nuvla application management platform enables you to extend your DevOps workflow to the edge.

Benefits of DevOps

DevOps is not a new concept. Far from it. For a little over ten years now, this endearing set of best practices has been making its way into every software development and IT operations team worthy of its name.

In fact, a DevOps report from 2019 studied and gathered statistics on how this set of Agile practices makes a difference between teams who adopt it and those who don’t:

devops stats

In short, teams who adopt DevOps are in general more productive, deliver more reliable products, and do all of that much faster.

Continuous Integration & Delivery is key

One of the biggest DevOps tactics is called CI/CD, which stands for Continuous Integration and Continuous Delivery/Deployment. Its purpose is to integrate a trustworthy and automated chain of testing marks into the Agile development process.

For the sake of simplicity, let’s take the following DevOps workflow:


The Continuous Integration phase consists of everything from planning to coding, building and testing (pretty much the left part of the diagram above). This is where all the automated tests are executed against every new code change, and where your software is compiled and/or built.

Continuous Delivery/Deployment comes after that, by releasing the committed software, and in some cases, deploying it into a QA environment for validation (for example for Canary deployments).

What does Nuvla have to do with all of this?

Being able to continuously deploy your software across a fleet of edge devices sounds ideal. However, most software houses lack the ability to consistently deploy software on edge computing infrastructures. They typically have a large backlog of updates waiting to be manually deployed. This is where Nuvla can help. To get started with Nuvla you only need to:

Our documentation will support you along the way. And if you hit a snag, simply reach out to us via the little blue widget on the bottom right of any of our websites.

Developers, read on ⬇️

Well, DevOps without tooling would be quite complicated. You’ll easily find a plethora of tools for each step of the DevOps process: Issue Trackers and Scrum Boards for planning (like JIRA), IDEs for coding (like IntelliJ), Source Code Management systems for tracking changes (like GitHub), Continuous Testing applications for testing code changes (like Jenkins and GitHub CI), Continuous Deployment tools for automatic provisioning (like ArgoCD), etc.

So, if you think about it, since Nuvla is basically an application management platform with orchestration capabilities for Docker and Kubernetes environments, why not use it for Continuous Deployment?


So let’s say you plan for a new feature. An awesome one by the way. First, you develop it. Your code looks neat and tidy, so you commit it. When you commit, your continuous tests will be triggered, and they’ll run all the unit and integration tests to validate that your feature is working properly. And it is, so congrats! Upon this successful step, you will also have another action that will build and push your new code as an artefact (let’s say your final artefact is a Docker Image in Docker Hub). Finally, you want to roll out this new feature to all of your ongoing deployments (be it QA or production). So instead of doing this roll-out manually, why not simply trigger a new continuous action, that will use Nuvla’s REST API to deploy your new application (that is registered in your Nuvla App Store) into your infrastructures (also registered in Nuvla)?

This will give you transparent and infrastructure-agnostic deployments in any compute infrastructure (at the edge or in the cloud), plus you’ll be able to then operate and monitor the lifecycle of your deployments from the Nuvla platform!

Example of CD using Nuvla

To help you integrate Nuvla into your CI/CD pipeline, we at SixSq have built a GitHub Action that you can use to automatically deploy an app from Nuvla into an infrastructure.

This GitHub Action is called Nuvla Application Deployer and is available in the GitHub marketplace at

To use it, simply make sure your GitHub repositories has the GitHub Actions enabled. In your GitHub Actions Workflow, you can use our Nuvla Application Deployer as a job. Something like this:

Jobs: continuous-deployment: 

runs-on: ubuntu-latest
name: Deploy application from Nuvla steps: 

- name: Deploy
id: deploy
uses: nuvla/nuvla-deploy-app-action@v1 with: 

api-key: $ api-secret: $ module-id: 'module/<uuid>'
credential-id: 'credential/<uuid>' environment: 'NUVLABOX_UUID=nuvlabox/<uuid>' 

Use the output

run: echo “The output was $”

Looks quite straightforward, right? Just to make sure it’s all clear, let’s go through each action input:

If the action fails, it will try to auto-clean whatever deployments have been started, otherwise, on success, it will output the ID of your deployment, which you can capture via the workflow variable:


Now you can manage the lifecycle of your edge applications much more easily.

Continuous Deployment is just one of the many benefits of using Nuvla. There are several ways you can dig deeper, so why not take a look then contact us if you need a helping hand? And we are always interested to hear about your experience. So don’t hesitate to get in touch.

Latest blog posts

15 May 2024

Bouquet Of Apps At The Edge

By Marc-Elian Bégin

Why settle for something less than exquisite when you’re running artificial intelligence (AI) at the edge? We think you deserve a beautiful bouquet...

Read blog

13 May 2024

Running AI at the edge? It’s not rocket science!

By Marc-Elian Bégin

Running artificial intelligence (AI) and machine learning (ML) at the edge to turn your data into actionable insights isn’t rocket science. I shoul...

Read blog

22 March 2024

Enhancing AI strategies with CPE – a telco perspective

By Marc-Elian Bégin

If I tell you the internal temperature of a specific outdoor telecom cabinet has been reported as 55⁰C (131⁰F) multiple times, would that concern y...

Read blog