Why Pods are the smallest deployable unit in Kubernetes - Why It Works

Introduction

In Kubernetes, a Pod is the smallest unit you can deploy and manage. It groups one or more containers that share resources and run together on the same machine. This helps keep related containers close and makes managing them easier.

When you want to run a single container or a tightly coupled group of containers together on the same host.

When you need containers to share storage or network resources easily.

When you want to manage the lifecycle of related containers as a single unit.

When you want to ensure containers start, stop, and scale together.

When you want to simplify deployment by grouping containers that must work closely.

Commands

This command creates a Pod named 'my-pod' running the nginx container. The '--restart=Never' flag ensures it creates a Pod directly, not a Deployment.

Terminal

kubectl run my-pod --image=nginx --restart=Never

Expected Output

pod/my-pod created

→

--restart=Never - Creates a Pod directly instead of a Deployment

This command lists all Pods in the current namespace to verify that 'my-pod' is running.

Terminal

kubectl get pods

Expected Output

NAME     READY   STATUS    RESTARTS   AGE

my-pod   1/1     Running   0          10s

This command shows detailed information about the 'my-pod' Pod, including its containers, status, and resource usage.

Terminal

kubectl describe pod my-pod

Expected Output

Name:         my-pod

Namespace:    default

Priority:     0

Node:         <node-name>

Start Time:   <time>

Labels:       <none>

Annotations:  <none>

Status:       Running

IP:           <pod-ip>

Containers:

  nginx:

    Container ID:   <container-id>

    Image:          nginx

    State:          Running

    Ready:          True

    Restart Count:  0

Conditions:

  Type           Status

  Initialized    True 

  Ready          True 

  ContainersReady True 

  PodScheduled   True 

Volumes:

  <none>

QoS Class:       BestEffort

Node-Selectors:  <none>

Tolerations:     <none>

Events:          <none>

Key Concept

A Pod is the smallest deployable unit because it groups containers that share resources and lifecycle, making them manageable as one.

Common Mistakes

Trying to deploy individual containers without using Pods

Containers need to be inside Pods to run in Kubernetes; you cannot deploy containers alone.

Always create Pods to run containers, either directly or via higher-level objects like Deployments.

Using 'kubectl run' without '--restart=Never' expecting a Pod but getting a Deployment

'kubectl run' defaults to creating a Deployment unless '--restart=Never' is specified.

Add '--restart=Never' to create a Pod directly with 'kubectl run'.

Summary

Pods are the smallest deployable units in Kubernetes that hold one or more containers.

Pods share storage, network, and lifecycle, making container management simpler.

Use 'kubectl run' with '--restart=Never' to create a Pod directly and 'kubectl get pods' to verify.

Practice

1. Why are Pods considered the smallest deployable unit in Kubernetes?

easy

A. Because they group one or more containers that share resources like network and storage

B. Because they represent a single container only

C. Because they are the smallest unit of storage in Kubernetes

D. Because they are the smallest unit of CPU allocation

Solution

1. Step 1: Understand what a Pod contains

   A Pod groups one or more containers that share the same network and storage resources.

2. Step 2: Identify why this grouping matters

   This grouping allows containers to be managed and deployed together as a single unit.

3. Final Answer:

   Because they group one or more containers that share resources like network and storage -> Option A

4. Quick Check:

   Pod = smallest deployable unit grouping containers [OK]

Hint: Pods group containers sharing resources, making them smallest units [OK]

Common Mistakes:

• Thinking Pods are single containers only
• Confusing Pods with storage units
• Assuming Pods allocate CPU individually

2. Which of the following is the correct way to define a Pod in a Kubernetes YAML file?

easy

A. apiVersion: v1\nkind: Service\nmetadata:\n name: mypod\nspec:\n containers:\n - name: mycontainer\n image: nginx

B. apiVersion: v1\nkind: Pod\nmetadata:\n name: mypod\nspec:\n containers:\n - name: mycontainer\n image: nginx

C. apiVersion: apps/v1\nkind: Pod\nmetadata:\n name: mypod\nspec:\n containers:\n - name: mycontainer\n image: nginx

D. apiVersion: v1\nkind: Pod\nmetadata:\n name: mypod\nspec:\n image: nginx

Solution

1. Step 1: Check the apiVersion and kind fields

   The correct apiVersion for a Pod is 'v1' and kind is 'Pod'.

2. Step 2: Verify the containers section

   The containers must be listed under spec with name and image fields inside a list.

3. Final Answer:

   apiVersion: v1\nkind: Pod\nmetadata:\n name: mypod\nspec:\n containers:\n - name: mycontainer\n image: nginx -> Option B

4. Quick Check:

   Pod YAML needs apiVersion v1 and kind Pod [OK]

Hint: Pod YAML uses apiVersion v1 and kind Pod with containers list [OK]

Common Mistakes:

• Using wrong kind like Service instead of Pod
• Missing containers list under spec
• Wrong apiVersion for Pod

3. Given this Pod spec snippet, what will happen when it is deployed?

apiVersion: v1
kind: Pod
metadata:
  name: testpod
spec:
  containers:
  - name: app
    image: busybox
    command: ['sleep', '3600']
  - name: sidecar
    image: busybox
    command: ['sleep', '3600']

medium

A. Containers will run but on separate Pods automatically

B. Only the first container will run, the second will be ignored

C. The Pod will fail because multiple containers are not allowed

D. Two containers will run together sharing the same network and storage

Solution

1. Step 1: Analyze the Pod spec with multiple containers

   The spec defines two containers inside one Pod, which is allowed and common.

2. Step 2: Understand container behavior in a Pod

   Both containers run together sharing the Pod's network and storage resources.

3. Final Answer:

   Two containers will run together sharing the same network and storage -> Option D

4. Quick Check:

   Multiple containers in one Pod run together sharing resources [OK]

Hint: Multiple containers in one Pod run together sharing network/storage [OK]

Common Mistakes:

• Thinking only one container runs per Pod
• Assuming multiple containers cause Pod failure
• Believing containers run in separate Pods automatically

4. You deployed a Pod with two containers, but only one container is running. What is the most likely cause?

medium

A. The Pod spec must list containers under metadata, not spec

B. Pods can only run one container at a time

C. The second container's image name is incorrect causing it to fail

D. Kubernetes does not support multiple containers in a Pod

Solution

1. Step 1: Recall Pod container rules

   Pods can run multiple containers simultaneously if defined correctly.

2. Step 2: Identify common failure reasons

   If one container is not running, a common cause is an incorrect image name or failure to pull the image.

3. Final Answer:

   The second container's image name is incorrect causing it to fail -> Option C

4. Quick Check:

   Container failure often due to wrong image name [OK]

Hint: Check container image names if some containers don't run [OK]

Common Mistakes:

• Believing Pods run only one container
• Misplacing containers under metadata
• Thinking Kubernetes disallows multiple containers in Pods

5. You want to deploy two tightly coupled containers that must share the same IP address and storage. Why should you deploy them in a single Pod instead of separate Pods?

hard

A. Because containers in the same Pod share network and storage, enabling tight coupling

B. Because separate Pods cannot run containers at the same time

C. Because Kubernetes does not allow multiple Pods to communicate

D. Because Pods automatically merge containers into one process

Solution

1. Step 1: Understand resource sharing in Pods

   Containers in the same Pod share the same IP address and storage volumes.

2. Step 2: Compare with separate Pods

   Separate Pods have different IPs and isolated storage, making tight coupling harder.

3. Step 3: Reason why single Pod is best

   Deploying tightly coupled containers in one Pod ensures they can communicate easily and share data.

4. Final Answer:

   Because containers in the same Pod share network and storage, enabling tight coupling -> Option A

5. Quick Check:

   Shared network/storage in Pod enables tight container coupling [OK]

Hint: Use one Pod to share IP and storage for tightly coupled containers [OK]

Common Mistakes:

• Thinking separate Pods can't run simultaneously
• Believing Kubernetes blocks Pod communication
• Assuming Pods merge containers into one process