Skip to main content
Version: Cloud

Install the Deeploy software stack

This repository contains detailed instructions for installing the Deeploy Core software stack. Here, you will find all the necessary scripts to configure the dependencies (Istio, Knative, and cert-manager), as well as the installation process for the Deeploy Core Helm chart.

For installation of the chart, see https://artifacthub.io/packages/helm/deeploy-core/deeploy and modify the values file as described below.

Azure values

Alter the necessary values for your Helm installation. You can access example values files locally by running the following command: helm pull deeploy-install/deeploy --version 1.37.1 --untar.

OpenID Connect values (optional)

To set up OpenID Connect, follow these steps:

  • Create a new App Registration
  • Use the generated application ID as your clientId
  • Within the App Registration, go to manage authentication add a platform
  • Add a web platform and use https://api.<HOST>/auth/self-service/methods/oidc/callback/microsoft as the allowed redirect URI
  • Within the App Registration, go to manage certificates & secrets new client secret
  • Use this generated secret as your clientSecret

To restrict or allow access to users, you could:

  • Go to the generated Enterprise Application
  • Manage users in the users and groups tab

Upgrades

Make sure to regularly upgrade your Deeploy installation. Specific instructions are listed in the release notes (tags). The upgrade process consists of two steps:

  1. Infrastructure update (e.g., upgrade EKS Kubernetes version)
  2. Switch to a new version of Deeploy:
    • Patch (e.g., 1.34.0 to 1.34.1), small changes often a result of a bug fix
    • Major release (e.g., 1.33.0 to 1.34.0), larger (visual) changes that can include infrastructure changes

MyDeeploy license registration

When you install Deeploy for the first time, the Deeploy helm chart will register your license key (runPreInstall: true). For Azure Marketplace installations, registering a license through MyDeeploy is not necessary.