22. Controller - DaemonSet

Table of contents

22. Controller - DaemonSet

Summary

A DaemonSet is a controller that deploys pods to all nodes or a specific node group in a cluster.
Unlike other replication controllers, DaemonSets cannot be scheduled for failure - they must always run on every node.
When new nodes are added, DaemonSets automatically deploy pods to those nodes.
When nodes are removed, DaemonSets automatically delete the pods running on those nodes.
DaemonSets are commonly used for deploying system processes such as log collectors or monitoring agents, which need to run on every node in the cluster.

DaemonSet definition

ReplicaSet

        
apiVersion: apps/v1     #rc apiVersion
kind: ReplicaSet        #rc kind
metadata:
  name: rs-nginx        #rc name
spec:
  replicas: 3           # Min. Pods Numb.
  selector:
    matchLabels:
      app: webui        #no meaning
  template:
    metadata:
      name: nginx-pod
      labels:
        app: webui
    spec:
    containers:
    - name: nginx-container     #container name
            image: nginx:1.14   #container image

Deployment

        
apiVersion: apps/v1                # DaemonSet apiVersion
kind: DaemonSet                    # DaemonSet kind option
metadata:
  name: daemonset-nginx
spec:
  selector:                        # no replicas option
    matchLabels:
      app: webui
  template:
    metadata:
      name: nginx-pod
      labels:
        app: webui
    spec:
      containers:
      - name: nginx-container
        image: nginx:1.14

DaemonSet example-1

Prepare test

Create DaemonSet by yaml

vim daemonset-exam.yaml

apiVersion: apps/v1
kind: DaemonSet
metadata:
  name: daemonset-nginx
spec:
  selector:
    matchLabels:
      app: webui
  template:
    metadata:
      name: nginx-pod 
      labels:
        app: webui
    spec:
      containers:
      - name: nginx-container
        image: nginx:1.14

Turn on the watch mode for pods(T1)
```
watch kubectl get pods -o wide
```
Delete node3
```
kubectl delete nodes node3
```
enforce daemonset-exam.yaml
```
kubectl create -f daemonset-exam.yaml
```
Check the daemonset status again
```
kubectl get daemonset
```
Check the T1 watch again

test

Check the token list at Master
```
sudo kubeadm token list
```
Get a new token reissued.
```
sudo kubeadm token create --ttl 1h
```

Get a new hash code

sudo openssl x509 -pubkey -in /etc/kubernetes/pki/ca.crt | openssl rsa -pubin -outform der 2>/dev/null | openssl dgst -sha256 -hex | sed 's/^.* //'

Node3 re-register to Master

sudo kubeadm join (Master Ip Address):6443 --token (Token value) --discovery-token-ca-cert-hash sha256:
(Token hash code)

Check the DaemonSet status at Master

T1(watch kubectl get pods -o wide)
deamonset status
```
kubectl get daemonset.apps
```

DaemonSet example-2: Rolling update & Rollback

Check the pods and daemonset status

Turn on watching mode to pods
Turn on watching mode to daemonset

Test

Change Nginx version to 1.15 in edit mode

kubectl edit daemonsets.app daemonset-nginx

...
      spec:
        containers:
        - image: nginx:1.14     # Change to nginx:1.15
          imagePullPolicy: IfNotPresent
        name: nginx-container
        resources: {}
        terminationMessagePath: /dev/termination-log
        terminationMessagePolicy: File
...

<!– <div class="video-container">

</div>

–>

Check the pods event

Pod in node3
…
Pod in node2 …
Pod in node3 …

DaemonSet example-3: Rolling Rollback

Check the image version check within pod’s container
```
kubectl get daemonset.apps -o wide
```
Downgrade container image nginx to before current
```bash kubectl rollout undo daemonset daemons