Container Network Interface (CNI) in Kubernetes - Step-by-Step Execution

Process Flow - Container Network Interface (CNI)

Pod requests network

↓

Kubelet calls CNI plugin

↓

CNI plugin configures network

↓

Assign IP & setup routes

↓

Pod network ready

↓

Pod communicates over network

Shows how Kubernetes uses CNI plugins to set up networking for pods step-by-step.

Execution Sample

Kubernetes

kubectl apply -f pod.yaml
# Pod starts
# Kubelet calls CNI plugin
# CNI assigns IP
# Pod network ready

This sequence shows how a pod's network is set up using CNI when the pod starts.

Process Table

Step	Action	Component	Result	Notes
1	Pod creation requested	User/Kubernetes API	Pod object created	Pod spec includes network request
2	Kubelet detects pod start	Kubelet	Calls CNI plugin	Triggers network setup
3	CNI plugin invoked	CNI plugin	Reads config and allocates IP	Allocates IP from network pool
4	Network interfaces configured	CNI plugin	Network namespace setup	Routes and interfaces added
5	Pod network ready	Kubelet	Pod can send/receive packets	Pod joins cluster network
6	Pod communicates	Pod	Network traffic flows	Pod uses assigned IP
7	Pod deleted	User/Kubernetes API	Kubelet calls CNI DEL	Network resources cleaned up
8	CNI cleans network	CNI plugin	IP released and interfaces removed	Prepares for next pod
9	End	System	Pod network lifecycle complete	Network resources freed

💡 Pod deleted and network cleaned up, ending the network lifecycle

Status Tracker

Variable	Start	After Step 3	After Step 4	After Step 5	Final
Pod IP	None	10.244.1.5	10.244.1.5	10.244.1.5	None
Network Interfaces	None	eth0 created	eth0 configured	eth0 active	Removed
Routing Table	Empty	Routes added	Routes active	Routes active	Routes removed

Key Moments - 3 Insights

Why does the pod have no IP before the CNI plugin runs?

Before step 3 in the execution_table, the pod network is not set up, so no IP is assigned yet.

What happens if the CNI plugin fails to assign an IP?

The pod network setup fails at step 3, so the pod cannot communicate and will not become ready.

Why is the network cleaned up after pod deletion?

At steps 7 and 8, the CNI plugin removes interfaces and releases IPs to avoid resource leaks for future pods.

Visual Quiz - 3 Questions

Look at the execution_table, what IP is assigned to the pod after step 3?

A 10.244.1.5

B None

C 192.168.0.1

D 127.0.0.1

💡 Hint

Check the 'Pod IP' variable in variable_tracker after step 3

At which step does the pod network become ready for communication?

A Step 4

B Step 5

C Step 2

D Step 7

💡 Hint

Look for 'Pod network ready' in the execution_table

If the pod is deleted, what happens to the network interfaces?

A They remain active

B They are duplicated

C They are removed

D They are renamed

💡 Hint

See steps 7 and 8 in execution_table and variable_tracker final state

Concept Snapshot

Container Network Interface (CNI) in Kubernetes:
- Kubelet calls CNI plugin to set up pod network
- CNI plugin assigns IP and configures interfaces
- Pod network ready after setup
- On pod deletion, CNI cleans network resources
- Enables pod communication inside cluster

Full Transcript

Container Network Interface (CNI) is how Kubernetes sets up networking for pods. When a pod starts, the kubelet calls the CNI plugin. The plugin assigns an IP address and configures network interfaces and routes inside the pod's network namespace. Once done, the pod can communicate over the cluster network. When the pod is deleted, the kubelet calls the CNI plugin to clean up network resources like IP addresses and interfaces. This process ensures pods have isolated, functional networking and resources are reused efficiently.

Practice

1. What is the main purpose of the Container Network Interface (CNI) in Kubernetes?

easy

A. To manage storage volumes for containers

B. To provide IP addresses and networking for pods

C. To schedule pods on nodes

D. To monitor container resource usage

Solution

1. Step 1: Understand CNI role in Kubernetes

   CNI is responsible for setting up networking so pods can communicate.

2. Step 2: Identify correct function

   Providing IP addresses and network connectivity to pods is the core function of CNI.

3. Final Answer:

   To provide IP addresses and networking for pods -> Option B

4. Quick Check:

   CNI = Pod networking [OK]

Hint: CNI is about pod networking, not storage or scheduling [OK]

Common Mistakes:

• Confusing CNI with storage management
• Thinking CNI schedules pods
• Assuming CNI monitors resources

2. Which of the following is a valid CNI configuration file format?

easy

A. XML

B. CSV

C. INI

D. JSON

Solution

1. Step 1: Recall CNI config formats

   CNI plugins use JSON or YAML files to configure networking.

2. Step 2: Identify valid format

   JSON is the standard format for CNI configuration files.

3. Final Answer:

   JSON -> Option D

4. Quick Check:

   CNI config = JSON [OK]

Hint: CNI configs are JSON or YAML, not XML or INI [OK]

Common Mistakes:

• Choosing XML as config format
• Confusing INI or CSV with CNI config
• Not knowing CNI uses JSON

3. Given this CNI config snippet:

{
  "cniVersion": "0.4.0",
  "name": "mynet",
  "type": "bridge",
  "bridge": "cni0",
  "ipMasq": true,
  "ipam": {
    "type": "host-local",
    "subnet": "10.22.0.0/16"
  }
}

What IP subnet will pods get addresses from?

medium

A. 172.16.0.0/12

B. 192.168.1.0/24

C. 10.22.0.0/16

D. None, IPs are assigned dynamically

Solution

1. Step 1: Locate subnet in config

   The subnet is specified inside the "ipam" section as "10.22.0.0/16".

2. Step 2: Understand IP assignment

   Pods will get IPs from the defined subnet range in the config.

3. Final Answer:

   10.22.0.0/16 -> Option C

4. Quick Check:

   Subnet in ipam = 10.22.0.0/16 [OK]

Hint: Look inside "ipam" for subnet to find pod IP range [OK]

Common Mistakes:

• Ignoring the subnet field
• Choosing default Kubernetes pod subnet
• Thinking IPs are random without subnet

4. You installed a CNI plugin but pods cannot communicate across nodes. What is a likely cause?

medium

A. CNI config file missing or incorrect

B. Pods have no CPU limits set

C. Kubernetes API server is down

D. Node disk space is full

Solution

1. Step 1: Analyze pod network failure

   Pods not communicating across nodes usually means network setup is broken.

2. Step 2: Check CNI config

   If the CNI config file is missing or wrong, networking won't work properly.

3. Final Answer:

   CNI config file missing or incorrect -> Option A

4. Quick Check:

   Network issues = CNI config problem [OK]

Hint: Check CNI config file first for network issues [OK]

Common Mistakes:

• Blaming CPU limits for network problems
• Assuming API server affects pod networking directly
• Ignoring CNI config correctness

5. You want to deploy a new CNI plugin that supports network policies and IP address management. Which step is NOT required?

hard

A. Restart the Kubernetes API server

B. Create a CNI configuration file in /etc/cni/net.d/

C. Install the CNI plugin binaries on all nodes

D. Ensure kubelet is configured to use the new CNI

Solution

1. Step 1: Understand CNI plugin deployment steps

   You must install plugin binaries, add config files, and configure kubelet to use it.

2. Step 2: Identify unnecessary step

   Restarting the Kubernetes API server is not needed for CNI plugin changes.

3. Final Answer:

   Restart the Kubernetes API server -> Option A

4. Quick Check:

   API server restart NOT needed for CNI install [OK]

Hint: API server restart not needed when changing CNI [OK]

Common Mistakes:

• Thinking API server restart is required
• Forgetting to install plugin binaries
• Not placing config in correct directory