Overview
One of the challenges Cloud Providers face is that end users using their service are generally unfamiliar with the nuances of Kubernetes. They prefer to provide their users with an App Store type experience for all applications where the user just has to select/click the app in the portal to deploy and use the application.
In this exercise, you will take a custom app and learn how you can encapsulate it using Rafay's PaaS Studio and Environment Manager to deliver a 1-click consumption experience for end users. The illustrative image below shows an example of the streamlined, app-store style end user experience that can be achieved with this.
Info
In our example, we will deploy a simple webserver custom app onto a Kubernetes cluster. The web server will print "Hello < Input_Name >", where Input_Name is an Input provided by the user deploying the application.
Assumptions¶
This exercise assumes that you have access to the following:
- Access to a Rafay Org with Org Admin privileges
- Docker installed on your machine (Windows or Linux or macOS)
- An account on GitHub to perform GitOps based lifecycle operations
Considerations¶
For real life deployments, for web-based, custom applications, administrators will have to consider the following items. We will be skipping these in this exercise to keep things simple.
1. Ingress¶
The web-based application will need to be exposed via Kubernetes Ingress. As part of deployment, the Ingress needs to be created.
2. DNS/Domain¶
End users will access the web-based application using a web-browser by clicking on a URL. As part of deployment, the DNS entry needs to be auto programmed.
3. HTTPS Certificate¶
End users will access the web-based application using a web-browser by clicking on a https secured URL. As part of deployment, a certificate needs to be issued/assigned.
Info
All the steps described above are automatically performed for Rafay curated AI/ML web-based applications such as Jupyter Notebooks, Ray and Kubeflow based MLOps.