6. Labels and Annotations

Overview

Metadata is extremely useful for any organization and has its use in managing potentially thousands of resources in a cluster. This chapter teaches you how to add metadata to your pods or any other Kubernetes objects. You will be introduced to the concept of labels and annotations. We will also explain their use cases so that you can decide whether to use labels or annotations for a particular use case. You'll utilize labels to organize your objects by using label selectors to select or filter organized sets of objects. You'll also use annotations to add unstructured metadata information to objects.

Introduction

In the previous chapter, we created various kinds of pods and managed their life cycles. Once we start working with different pods, ideally, we would want to organize, group, and filter them based on certain properties. To do that, we need to add some information to our pods so that we can later use that information to organize them. We have already seen the use of the name and namespace fields as metadata for the pods. In addition to those fields, we can also add key-value pairs to the pods in order to add extra information as labels and annotations.

In this chapter, we will assign metadata to these pods in order to identify the pods through queries based on some metadata and then add additional unstructured metadata. We will cover labels and annotations in detail and examine the differences between them. We will use both labels and annotations and see when to use one or the other.

Labels

Labels are the metadata that contain identifiable information pertaining to the Kubernetes objects. These are basically key-value pairs that can be attached to objects such as pods. Each key must be unique for an object. Labels contain information that is meaningful to users. Labels can be attached to pods at the time of creation and can also be added or modified during their runtime too. Here is an example of how labels in a YAML file would appear:

metadata:
  labels:
    key1: value1
    key2: value2

Constraints for Labels

As noted earlier, labels are key-value pairs. There are certain rules that label keys and values should follow. These constraints exist because this way, the queries using labels can be evaluated faster by using optimized data structures and algorithms internally. Kubernetes internally maintains the mappings of labels to corresponding objects using optimized data structures to make these queries...

Annotations

As we have seen previously, labels are used to add the identifying metadata that we can later use to filter or select objects by. However, labels have certain constraints in terms of what we can store in the values, such as the limitation of 63 characters and alphanumeric characters at the beginning and end. Annotations, on the other hand, have fewer constraints in terms of what kind of data can be stored in them. However, we cannot filter or select objects by using annotations.

Annotations are also key-value pairs that can be used to store the unstructured information pertaining to the Kubernetes objects. Here is an example of how annotations in a YAML file would appear:

metadata:
  annotations:
    key1: value1
    key2: value2

Constraints for Annotations

As noted in the previous section, annotations are key-value pairs, just like labels. However, the rules for annotations are more relaxed than the rules for...

Summary

In this chapter, we have described labels and annotations and used them to add metadata information, which can either be identifiable information that can be used to filter or select objects, or non-identifiable information that can be used by users or tools to get more context regarding the state of the application. More specifically, we have also organized objects such as pods using labels and annotations. These are important skills that will help you manage your Kubernetes objects more efficiently.

In the following chapters, as we become familiar with more Kubernetes objects such as Deployments and Services, we will see the further application of labels and label selectors while organizing pods for deployment or discovery.