Whereas many firms are writing brand-new Kubernetes-based purposes, it’s nonetheless fairly frequent to seek out firms who wish to migrate current workloads. A typical supply platform for these purposes is Cloud Foundry. Nevertheless, getting an current Cloud Foundry software operating on Kubernetes may be non-trivial, particularly if you wish to keep away from making code modifications in your purposes, or taking over massive course of modifications throughout groups. That’s, in case you’re not utilizing Kf to do lots of that heavy lifting for you.
Kf is a Google Cloud service that permits you to simply transfer current Cloud Foundry workloads to Kubernetes with minimal disruption to your current processes.
Kf encompasses a command line interface (CLI) additionally named kf, that replaces the present Cloud Foundry cf command line utility. The kf CLI implements probably the most generally used cf performance, together with the flexibility to handle bindings, providers, apps, routes and extra.
For instance, to deploy an current software you’ll merely situation the kf push command.
On the server aspect Kf is constructed on a number of open supply applied sciences. In some circumstances these applied sciences are additionally the Google Cloud implementation. For example GKE is our managed Kubernetes providing, and offers the platform for managing and operating the purposes. Routing and ingress is dealt with by Anthos Service Mesh, Google Cloud’s managed Istio-based service mesh. Lastly, Tekton offers on-cluster construct performance for Kf. Builders haven’t got to fret about any of these applied sciences, as Kf abstracts them away.
Kf primitives corresponding to areas, bindings and providers are applied as customized Kubernetes assets and controllers. The customized assets successfully function the Kf API and are utilized by the kf CLI to work together with the system. The controllers use Kf’s CRDs to orchestrate the opposite parts within the system.
The fantastic thing about this method is that builders who’re accustomed to current workflows can largely replicate these workflows with the kf CLI. However, platform operators who’re extra accustomed to Kubernetes can use kubectl to work together with the CRDs and controllers.
For example in case you needed to record the apps operating in your Kf cluster you can situation both of the next instructions:
Discover that CF / Kf areas get mapped one to 1 to Kubernetes namespaces.
To get an inventory of all of the customized assets you’ll be able to study the api-resources within the kf.dev API group.
With Kf builders can proceed to work with a well-known interface and platform operators can use declarative Kubernetes practices and tooling corresponding to Anthos Config Administration to handle the cluster. It’s actually the very best of each worlds in case you’re seeking to handle your current Cloud Foundry purposes on Kubernetes.
For those who’d prefer to be taught extra about Kf take a look at the video I simply launched on YouTube. It opinions a few of the ideas mentioned right here, and features a quick demo. For those who’d prefer to get palms on, attempt the fast begin. And, in fact, you’ll be able to at all times learn the documentation.