A Quick Introduction to Istio, Prometheus, and Flagger
This lesson gives an introduction to Istio, Prometheus and Flagger.
We'll cover the following
Istio#
Istio is a service mesh that runs on top of Kubernetes. Quickly after the project was created, it became one of the most commonly used in Kubernetes. It allows traffic management that enables us to control the flow of traffic and other advanced networking such as:
- point to point security
- policy enforcement
- automated tracing
- monitoring
- logging
We could write a full book about Istio. Instead, we’ll focus on the traffic shifting and metric gathering capabilities of Istio and how we can use those to enable Canary deployments.
Prometheus#
We can configure Istio to expose metrics that can be pulled by specialized tools. Prometheus is a natural choice, given that it is the most commonly used application for pulling, storing, and querying metrics.
Its format for defining metrics can be considered the de-facto standard in Kubernetes. It stores time-series data that can be queried using its query language PromQL. We just need to make sure that Istio and Prometheus are integrated.
Flagger#
Flagger is a project sponsored by Weaveworks. It can use service mesh solutions like Istio, Linkerd, Gloo (which we used with Knative), and quite a few others. Together with a service mesh, we can use Flagger to automate deployments and rollback using a few different strategies. Even though the focus right now is canary deployments, you should be able to adapt the examples that follow into other strategies as well. To make things easier, Flagger even offers a Grafana dashboard to monitor the deployment progress.
🔍 Please note that we could have used Gloo instead of Istio, just as we did in the Using Jenkins X To Define And Run Serverless Deployments chapter. But, I thought that this would be an excellent opportunity to introduce you to Istio. Also, you should be aware that none of the tools are the focus of the book and that the main goal is to show you one possible implementation of canary deployments. Once you understand the logic behind it, you should be able to switch to whichever toolset you prefer.
🔍 This course is dedicated to continuous delivery with Jenkins X. All the other tools we use are chosen mostly to demonstrate integrations with Jenkins X. We are not providing an in-depth analysis of those tools beyond their usefulness to continuous delivery.
In the next lesson, we will be taking a look at how to install Istio, Prometheus, and Flagger.
