Choose between embedded and existing cluster
Before installing the application, you'll have to choose between two installation types: Embedded and Existing Cluster. This guide will explain the differences between both methods and help you decide which is the best for your needs.
#What is an embedded cluster installation
The Embedded Installation is an "all in one" installation, where the Kubernetes cluster is created on your machine and all components needed for the application are installed.
You can find more details about Embedded Cluster Installation in the Replicated documentation.
#What is an existing cluster installation
You can install the application on an existing Kubernetes cluster that already contains nodes. You have to check that this cluster meets the system and network requirements.
You can find more details about Existing Cluster Installation in the Replicated documentation
#How to choose between these options
Before installing the application, you'll have to reply to some considerations:
- do you have the resources and knowledge to manage your Kubernetes cluster?
- do you have security policies concerning applications deployed by Kube?
If you already have a Kubernetes cluster, that is running different applications, with precise management and security policies and a dedicated maintainer in your teams, you should use the existing cluster installation.
If you don't have resources dedicated to the maintenance of a Kubernetes cluster, prefer the embedded installation.
#Using external databases
For both solutions, you can use databases external from your Kubernetes cluster. The database preparation section will help you correctly configure them for your environment.
If you are using external databases, remind to open ports between the GitGuardian application cluster and your databases.