Metadata and Traceability

With the adoption of independently deployable services-based architectures, the complexity of managing these components and their inter-relationships is becoming problematic. In the following sections, we will outline a number of techniques that can assist you with this.

Labels

As mentioned earlier, labels and label selectors are among the more powerful metadata management features of Kubernetes. At its core, labels are a collection of text-based key/value pairs that can be attached to one or more objects: Pods, services, Deployments, and so on. Labels are intended to add information/semantics to objects that are relevant to the user and not the core Kubernetes system. A label selector is a method by which a user can group items together that have the same labels.

One of the most common uses of labels and label selectors in Kubernetes is the way that services use label selectors to group related Pods as endpoints for the service.

It's probably...