Readme.md

Kubernetes Labels and Annotations

Introduction

The effectiveness of the Kubernetes API comes from how it manages the Kubernetes resources via metadata: labels and annotations. Metadata is essential for grouping resources, redirecting requests and managing deployments. In addition, is is also used to troubleshoot Kubernetes applications.

Use Case

Labels can be used to help manage release management of your applications/deployments into your clusters, as an example.

• Labels - used by Kubernetes to make decisions and identify resources, such as which pods to load balance incoming traffic to.
• Annotations - Used to store non-arbitary data for resources, such as contact details of who manages the resource. This data will not be used by Kubernetes.

Research

Blog - Best Practices Guide for Kubernetes Labels and Annotations

Kubernetes Docs - Labels and Selectors - Annotations - Common Labels - Well Know Labels, Annotations and Taints

If you want to take this further, I previously wrote a blog post looking at how we can use Kubernetes Admission Control Open Policy Agent (OPA) to ensure Labels are defined on resources to ensure our backup software protects the data.

Try yourself

Get the labels for a node.

1. This will output all of the information for this node, however at the top of this, you will see annotations and labels under the metadata heading.

kubectl get node {node_name} -o YAML

If you have jq installed, you can also get the information in the following way ````kubectl get node {node_name} -o json | jq .metadata.labels``` 2. To filter this down to just labels.

kubectl get node {node_name} --show-labels=true

Modifier explanation
--show-labels=false: When printing, show all labels as the last column (default hide labels column)

We can now get all our nodes which share the same label, for example

kubectl get nodes --selector beta.kubernetes.io/arch=amd64

Modifier explanation
-l, --selector='': Selector (label query) to filter on, supports '=', '==', and '!='.(e.g. -l key1=value1,key2=value2)

3. See how labels are used by controllers for selection criteria in Kubernetes. Services are a great example of using labels for selection criteria.

kubectl get svc {svc_name} -n {namespace} -o yaml

If you've deployed my pacman application into Kubernetes, you can see this with the Load Balancer service for the web front end

kubectl get svc pacman -n pacman -o yaml

I've concatenated my output response below

apiVersion: v1
kind: Service
metadata:
  creationTimestamp: "2021-06-03T17:41:29Z"
  finalizers:
  - service.kubernetes.io/load-balancer-cleanup
  labels:
    name: pacman
  managedFields:
.......
spec:
  clusterIP: 10.100.101.134
  clusterIPs:
  - 10.100.101.134
  externalTrafficPolicy: Cluster
  ipFamilies:
  - IPv4
  ipFamilyPolicy: SingleStack
  ports:
  - nodePort: 31033
    port: 80
    protocol: TCP
    targetPort: 8080
  selector:
    name: pacman
  sessionAffinity: None
  type: LoadBalancer

Create Labels on a node.

4. Adhoc add labels or annotations to an existing resource. From the Kubernetes Cheet Sheet.

kubectl label pods my-pod new-label=awesome                      # Add a Label
kubectl annotate pods my-pod icon-url=http://goo.gl/XXBTWq       # Add an annotation

Follow the rest of the blog post I linked to for a deeper dive and explanations of when and where to use Labels and Annotations.