Core Concepts
SUSE® Rancher Prime Continuous Delivery is fundamentally a set of Kubernetes custom resource definitions (CRDs) and controllers to manage GitOps for a single Kubernetes cluster or a large-scale deployment of Kubernetes clusters.
For more on the naming conventions of CRDs, click here. |
Below are some of the concepts of SUSE® Rancher Prime Continuous Delivery that will be useful throughout this documentation:
-
SUSE® Rancher Prime Continuous Delivery Manager: The centralized component that orchestrates the deployments of Kubernetes assets from git. In a multi-cluster setup, this will typically be a dedicated Kubernetes cluster. In a single cluster setup, the SUSE® Rancher Prime Continuous Delivery manager will be running on the same cluster you are managing with GitOps.
-
SUSE® Rancher Prime Continuous Delivery controller: The controller(s) running on the SUSE® Rancher Prime Continuous Delivery manager orchestrating GitOps. In practice, the SUSE® Rancher Prime Continuous Delivery manager and SUSE® Rancher Prime Continuous Delivery controllers are used fairly interchangeably.
-
Single Cluster Style: This is a style of installing SUSE® Rancher Prime Continuous Delivery in which the manager and downstream cluster are the same cluster. This is a very simple pattern to quickly get up and running with GitOps.
-
Multi Cluster Style: This is a style of running SUSE® Rancher Prime Continuous Delivery in which you have a central manager that manages a large number of downstream clusters.
-
SUSE® Rancher Prime Continuous Delivery agent: Every managed downstream cluster will run an agent that communicates back to the SUSE® Rancher Prime Continuous Delivery manager. This agent is just another set of Kubernetes controllers running in the downstream cluster.
-
GitRepo: Git repositories that are watched by SUSE® Rancher Prime Continuous Delivery are represented by the type
GitRepo
.
Example installation order via
GitRepo
custom resources when using SUSE® Rancher Prime Continuous Delivery for the configuration management of downstream clusters:
Install Calico CRDs and controllers.
Set one or multiple cluster-level global network policies.
Install GateKeeper. Note that cluster labels and overlays are critical features in SUSE® Rancher Prime Continuous Delivery as they determine which clusters will get each part of the bundle.
Set up and configure ingress and system daemons.
-
Bundle: An internal unit used for the orchestration of resources from git. When a
GitRepo
is scanned it will produce one or more bundles. Bundles are a collection of resources that get deployed to a cluster.Bundle
is the fundamental deployment unit used inSUSE® Rancher Prime Continuous Delivery. The contents of aBundle
may be Kubernetes manifests, Kustomize configuration, or Helm charts. Regardless of the source the contents are dynamically rendered into a Helm chart by the agent and installed into the downstream cluster as a helm release.-
To see the life cycle of a bundle, click here.
-
-
BundleDeployment: When a
Bundle
is deployed to a cluster an instance of aBundle
is called aBundleDeployment
. ABundleDeployment
represents the state of thatBundle
on a specific cluster with its cluster specific customizations. The SUSE® Rancher Prime Continuous Delivery agent is only aware ofBundleDeployment
resources that are created for the cluster the agent is managing.-
For an example of how to deploy Kubernetes manifests across clusters using SUSE® Rancher Prime Continuous Delivery customization, click here.
-
-
Downstream Cluster: Clusters to which SUSE® Rancher Prime Continuous Delivery deploys manifests are referred to as downstream clusters. In the single cluster use case, the SUSE® Rancher Prime Continuous Delivery manager Kubernetes cluster is both the manager and downstream cluster at the same time.
-
Cluster Registration Token: Tokens used by agents to register a new cluster.