A deployment of Kubernetes and the associated applications and services require the following components/prerequisites:
• A deployment machine - your local machine or a machine in the network to which you are deploying – either of which should be connected to the same network, physically or via VPN, as the cluster on which the deployment will occur. The deployment machine is used to provision the compnents of the fraXses platform required by the client. This machine should have the required tools installed.
• Access to Harbor – the Fraxses registry where all docker containers/images are stored. Login credentials to be provided by a Fraxses administrator.
• An existing cluster/group of machines/VM's on which the installation will occur OR a managed Kubernetes instance with a known number of nodes.
Lens is a Kubernetes client that provides a single of view of most (but not all) Kubernetes objects that will be required for the deployment. It provides an easy way to change yaml scripts, check container logs, scale deployments and execute into container as well as a raft of other things. We recommend it highly but if it isn't possible to use, or not your preference, then the same functionality can be had via kubectl (the kubernetes command line). Depending on client there may also be preferred Kubernetes clients in their environment (like Rancher's dashboard).
Docker is a ubiqutous container runtime. It is required to run the provisioning containers for implementation of the fraXses platform. Other container runtimes like rkt and frakti ARE NOT SUPPORTED by the platform team.
Harbor is the container registry provided by the platform team for delivery of the fraXses platform. Harbor can accessed via the web UI to view images and version. In order to pull container images for useage, you need to log into Harbor via the docker login command:
docker login registry.fraxses.com
Credentials will need to be provided in each case in order to login successfully.
When installing into a client environment (on-prem or managed), unless you are sitting in the clients network, you will need some form of VPN access or service account access to get to the Kubernetes cluster.