Working With ReplicaSets

Estimated time to read: 4 minutes

ReplicaSets in Kubernetes are a type of controller that ensures a specified number of identical Pod replicas are running at any given time. They help manage the availability of your applications by automatically creating or deleting Pods as necessary to match the desired number of replicas.

Key Concepts

Desired State: The number of Pod replicas that you want running in your cluster.
Current State: The number of Pod replicas currently running in your cluster.
Self-Healing: If a Pod fails or is deleted, the ReplicaSet controller automatically creates a new Pod to maintain the desired state.

Core Components of RS :

Specifying Replicas:
- The replicas field in the ReplicaSet manifest defines how many replicas of a pod should be running. For example, setting replicas: 4 ensures that four instances of the pod are running at all times.
- Pod Template:
- The template section of the ReplicaSet defines the pod specification. This includes container images, environment variables, ports, and other configurations.
- Label Selector:
- The selector section is used to match the labels on pods. It determines which pods the ReplicaSet is responsible for. This allows the ReplicaSet to manage specific pods based on their labels.
- Self-Healing:
- ReplicaSets are self-healing by design. If a pod dies or becomes unresponsive, the ReplicaSet replaces it automatically, ensuring the system remains in the desired state.

Example Configuration

Replicasets/ofl-replicaset.yml
apiVersion: apps/v1
kind: ReplicaSet
metadata:
  name: ofl-replicaset
spec:
  replicas: 4
  selector:
    matchLabels:
      environment: production
  template:
    metadata:
      labels: 
        environment: production
        tier: front-end
    spec:
      containers:
        - name: ofl-webserver
          image: nginx:latest
          ports:
          - containerPort: 80

Create ReplicaSet

kubectl apply -f Replicasets/ofl-replicaset.yml

List of the ReplicaSet
```
kubectl get rs
```

Describe ReplicaSet

kubectl describe replicaset ofl-replicaset

List out the all objects
```
kubectl get all
```

Scaling ReplicaSets

Scaling ReplicaSets in Kubernetes involves adjusting the number of pod replicas that are running at any given time. This can be done manually or automatically, depending on your needs.

Manual Scaling

You can manually scale a ReplicaSet by modifying the replicas field in the ReplicaSet YAML manifest or by using the kubectl command.

Scaling via kubectl Command
The easiest way to scale a ReplicaSet is to use the kubectl scale command. This command allows you to specify the desired number of replicas.
```
kubectl scale rs <rs-name> --replicas=5
```
```
kubectl scale rs ofl-replicaset --replicas=5
```

Scaling by Editing the YAML Manifest.

You can also scale a ReplicaSet by editing its YAML manifest directly and applying the changes.
Update the replicas field in the YAML file using kubectl command
```
kubectl edit rs ofl-replicaset
```

Update the replicas field in the YAML file.

Replicasets/ofl-replicaset.yml
apiVersion: apps/v1
kind: ReplicaSet
metadata:
  name: ofl-replicaset
spec:
  replicas: 4 # Change this number to the desired replica count
  selector:
    matchLabels:
      environment: production
  template:
    metadata:
      labels: 
        environment: production
        tier: front-end
    spec:
      containers:
        - name: ofl-webserver
          image: nginx:latest
          ports:
          - containerPort: 80

Apply the updated YAML file:

kubectl apply -f Replicasets/ofl-replicaset.yml

Update replica set using sed command

sed -i 's/4/6/g' Replicasets/ofl-replicaset.yml

Automatic Scaling

For dynamic scaling based on resource utilization (e.g., CPU, memory), you can use the Horizontal Pod Autoscaler (HPA). The HPA automatically scales the number of pods in a ReplicaSet (or Deployment) based on observed CPU utilization, memory usage, or custom metrics.

kubectl autoscale rs ofl-replicaset --min=2 --max=10 --cpu-percent=80

--min=2: Minimum number of replicas.
--max=10: Maximum number of replicas.
--cpu-percent=80: Target average CPU utilization across all pods.

You can check the status of the HPA using:

kubectl get hpa

The HPA will monitor the resource utilization of the pods managed by the ReplicaSet and adjust the number of replicas automatically within the specified range.

CleanUp

Delete ofl-replicaset

kubectl delete -f Replicasets/ofl-replicaset.yml

kubectl delete hpa ofl-replicaset