Continental Innovates with Rancher and Kubernetes
Today we are announcing support for Istio with Rancher 2.3 in Preview mode.
Istio, and service mesh generally, has developed a huge amount of excitement in the Kubernetes ecosystem. Istio promises to add fault tolerance, canary rollouts, A/B testing, monitoring and metrics, tracing and observability, and authentication and authorization, eliminating the need for developers to instrument or write specific code to enable these capabilities. In effect, developers can just focus on their business logic and leave the rest to Kubernetes and Istio.
The claims above aren’t new. About 10 years ago, PaaS vendors made exactly the same claim and even delivered on it to an extent. The problem was that their offerings required specific languages, frameworks, and, for the most part, only worked with very simple applications. The workloads were also tied to the vendor’s unique implementation, which meant that if you wanted your applications to use the PaaS services, you were potentially locked-in for a very long time.
With containers and Kubernetes, these limitations are virtually nonexistent. As long as you can containerize your application, Kubernetes can run it for you.
Our users count on us to make managing and operating Kubernetes and related tools and technologies easy, without locking them in to a specific cloud vendor. With Istio, we take the same approach.
In this Preview mode, we provide users with a simple UI to enable Istio under the Tools menu. Reasonable default configurations are provided but can be changed as required:
In order to monitor your traffic, Istio needs to inject an Envoy sidecar. In Rancher 2.3 Preview, users can enable automatic sidecar injection for each namespace. Once this option is selected, Rancher will inject the sidecar container into each workload:
Rancher’s simplified installation and configuration of Istio comes with a built-in, supported Kiali dashboard for traffic and telemetry visualization, Jaeger for tracing, and even its own Prometheus and Grafana (separate instances than the ones used for Advanced Monitoring).
After you deploy workloads in the namespaces with automatic sidecar injection enabled, head over to the Istio menu entry and observe the traffic as it flows across your microservice applications:
Clicking on Kiali, Jaeger, Prometheus, or Grafana will take you to the respective UI of each tool, where you can find more details and options:
As mentioned earlier, the power of Istio is its ability to bring features like fault tolerance, circuit breaking, canary deployment, and more to your services. To enable these, you will need to develop and apply the appropriate YAML files. Istio is not supported for Windows workloads yet, so it should not be enabled in Windows clusters.
Istio is one of the most talked about and requested features in the Rancher and Kubernetes communities today. However, there are also a lot of questions around the best way to deploy and manage it. With Rancher 2.3.0 Preview 2, our goal is to make this journey quick and easy.
For release notes and installation steps, please visit https://github.com/rancher/rancher/releases/tag/v2.3.0-alpha5