Approachable solutions to integration issues with Hub-X

hub-x, microservices architecture.

Hub-X is built according to the microservices architecture. It consists of connector services and customer services. The connector services are the components that retrieve information from or send information to a specific system. The customer services leverage the connector services and they ensure that the information retrieved by a connector service is transformed appropriately so that it can be sent by another connector service to a system that uses that information.

This allows us to connect new customers quickly and easily by building a customer service for each new customer that performs the desired transformations and uses existing or possibly new connector services. Customers can choose to run their own customer service in our shared platform, but for heavy workloads or high confidentiality requirements, they can also choose a dedicated solution.

use of new, innovative technology.

Whenever possible, we use the latest techniques and practices. Hub-X runs in Kubernetes and is built with C# and .NET core 6.0. We deploy with CI/CD pipelines from Azure Devops and use Git based source control. Sonar Qube is integrated into the pipelines for automated code quality assessment and we require at least 80% code coverage for unit tests. We review each other's pull requests and after review we deploy directly to the development environment.

Currently, our shared Hub-X platform is running in Azure Kubernetes Services. The intention is that in the future we can also offer Hub-X to customers as a dedicated solution and give them the choice to run in Azure, AWS or on their own infrastructure on Rancher. We will then define our own pipeline in Azure DevOps for each customer who wishes to do so.

Ricky's personal role and experience.

In this project, I was mainly involved in setting up the Kubernetes environment on Azure. I really enjoyed this, as I had been following the developments around Kubernetes and related technologies for a while. In the past, I had already given a presentation to colleagues on this topic in the context of knowledge sharing. For that, I had built a small sample application that was hosted in a local test cluster.

Fun, but too small to really learn all aspects of Kubernetes. Now I got a chance to use Kubernetes in practice and the specific problems I ran into in setting up a product worthy environment accelerated the depth of my knowledge.