Pod, Service, and Deployment

Pod, Service, and Deployment

Big Picture

Pods: Containers and Node Abstraction

What is a Pod?
  • A pod includes one or more containers plus additional share resources.
    • Shared resources include, but are not limited to, shared IP (internal/external) and storage volumes.
    • Example pod: one nginx webserver container plus one log-shipper sidecar container plus shared volume for log data.
  • Pods are scheduled onto nodes by the kube-scheduler.
Pod and physical (virtual) node
  • Each node runs a kubelet agent.
  • Kubelet talks to the container runtime (containerd in Rancher Desktop).
  • When a pod is deployed, kubelet pulls the associated container image(s) and runs it(them).
    graph TD
        subgraph Node1["Node (VM)"]
            Kubelet1["kubelet + runtime"]
            subgraph Pod1["Pod"]
                Nginx["Container: Nginx"]
            end
        end

        subgraph Node2["Node (VM)"]
            Kubelet2["kubelet + runtime"]
            subgraph Pod2["Pod"]
                Redis["Container: Redis"]
            end
        end

        Kubelet1 --> Pod1
        Kubelet2 --> Pod2
Hands-on with Rancher Desktop
  • Verify that your Rancher Desktop is up and running


1

 
kubectl get nodes -o wide
  • Create a file called nginx-pod.yaml with the following content


1
2
3
4
5
6
7
8
9
10
11
12

 
apiVersion: v1
kind: Pod
metadata:
    name: nginx
    labels:
        app: nginx
spec:
    containers:
    - name: nginx
      image: nginx:latest   # pulls Docker/OCI image
      ports:
      - containerPort: 80
  • Run kubectl and provide path to your nginx-pod.yaml. In the example below, I am in the same directory as my file.


1
2

 
kubectl apply -f nginx-pod.yaml
kubectl get pods -o wide

Services: Stable Access to Pods

At this point, if we try to access the above pod using the containerPort 80, it will fail.

Problem
  • Pods are ephemeral
    • Pods can restart and be rescheduled onto different nodes.
    • Each pod gets a random IP inside the cluster.
    • How does a client reliably connect to nginx if its internal IP changes?
  • Docker: -P and -p is not adequate for this.
Solution
  • Kubernetes Service
  • A Service provides Pods with a stable virtual IP and DNS name.
  • Service load-balances traffic to all matching Pods via label.
Connection to Physical Node
  • A ClusterIP service gives access only inside the cluster.
  • A NodePort service opens a port on every node’s IP.
  • A LoadBalancer (on cloud) provisions an external IP (if available).


1

 
kubectl get nodes
Hands-on with Rancher Desktop: Adding service to pod
  • Create a file called nginx-svc.yaml with the following content


1
2
3
4
5
6
7
8
9
10
11
12
13

 
apiVersion: v1
kind: Service
metadata:
    name: nginx
spec:
    type: NodePort
    selector:
        app: nginx
    ports:
    - protocol: TCP
      port: 80
      targetPort: 80
      nodePort: 30007

Deployment

Details
  • Pods creation using kubectl and Pods-only YAML files is a manual process.
    • Pod IP addresses will be ephemeral and changed when a Pod crashes/is deleted/is rescheduled.
  • Best practice:
    • Avoid creating bare Pods in production
    • Use Deployment (or StatefulSets, DaemonSets) to manage Pods
    • Combine with Service to maintain stable networking access.
Example: Details
  • Assuming that you have been working on this lecture continuously, you will have one nginx pod and one nginx service running. Use the following commmands to check the existence of the pod and service, then to delete the pod and service. After deletion, check again to confirm that the pod and service are gone.


1
2
3
4
5
6

 
kubectl get pods -o wide
kubectl get svc -o wide
kubectl delete pod nginx
kubectl delete svc nginx
kubectl get pods -o wide
kubectl get svc -o wide

 


1

 
<details class="details details--default" data-variant="default"><summary>Step 2: Create deployment</summary>

 


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

 
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
spec:
replicas: 2
selector:
    matchLabels:
    app: nginx
template:
    metadata:
    labels:
        app: nginx
    spec:
    containers:
    - name: nginx
      image: nginx:latest
      ports:
      - containerPort: 80
---
apiVersion: v1
kind: Service
metadata:
name: nginx-service
spec:
type: NodePort
selector:
    app: nginx
ports:
- protocol: TCP
  port: 80
  targetPort: 80
  nodePort: 30007

</details>

<summary>Step 3: Deployment</summary>

How does this work?
  • Deployment
    • Ensures 2 replics of nginx Pods always run
    • Each Pod get a random IP inside the cluster.
  • Service
    • Select all Pods with app: nginx.
    • Provide a stable virual IP and DNS name (nginx-service)
    • Expose port 30007 on every node.


1
2
3
4

 
kubectl apply -f nginx-deployment.yaml
kubectl get deployments
kubectl get pods -o wide
kubectl get svc

</details>

<summary>Step 4: Test recovery</summary>

</details>