Ingress vs LoadBalancer Service decision in Kubernetes - Performance Comparison

Time Complexity: Ingress vs LoadBalancer Service decision

O(n)

Understanding Time Complexity

When choosing between Ingress and LoadBalancer Service in Kubernetes, it's important to understand how their processing time grows as the number of requests or services increases.

We want to know how the system's work changes when more traffic or services are added.

Scenario Under Consideration

Analyze the time complexity of routing requests using an Ingress controller versus a LoadBalancer Service.

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: example-ingress
spec:
  rules:
  - host: example.com
    http:
      paths:
      - path: /app1
        pathType: Prefix
        backend:
          service:
            name: app1-service
            port:
              number: 80
---
apiVersion: v1
kind: Service
metadata:
  name: app1-loadbalancer
spec:
  type: LoadBalancer
  selector:
    app: app1
  ports:
  - port: 80
    targetPort: 8080

This snippet shows an Ingress routing multiple paths to services and a LoadBalancer Service exposing a single app directly.

Identify Repeating Operations

• Primary operation: Ingress controller checks each incoming request against its list of rules to find the matching path and service.
• How many times: For each request, it may check multiple rules until it finds a match.
• LoadBalancer Service: Directly routes requests to one service without rule checks.
• Dominant operation: Ingress rule matching per request.

How Execution Grows With Input

As the number of Ingress rules grows, the controller checks more rules per request, increasing work.

Number of Rules (n)	Approx. Checks per Request
10	Up to 10 checks
100	Up to 100 checks
1000	Up to 1000 checks

Pattern observation: The work grows roughly in direct proportion to the number of rules.

Final Time Complexity

Time Complexity: O(n)

This means the time to route a request grows linearly with the number of Ingress rules.

Common Mistake

[X] Wrong: "Ingress routing time stays the same no matter how many rules exist."

[OK] Correct: Each request must be checked against rules until a match is found, so more rules mean more checks and longer processing time.

Interview Connect

Understanding how routing scales helps you design systems that handle growth smoothly. This skill shows you can think about system behavior beyond just writing code.

Self-Check

"What if the Ingress controller used a hash map for rules instead of checking them one by one? How would the time complexity change?"

Practice

1. Which Kubernetes service type should you choose if you want to expose a single application with its own external IP address?

easy

A. LoadBalancer Service

B. Ingress Controller

C. ClusterIP Service

D. NodePort Service

Solution

1. Step 1: Understand service types

   LoadBalancer Service exposes an app with its own external IP, making it directly accessible.

2. Step 2: Compare with Ingress

   Ingress routes multiple apps through one IP, so it's not for single app direct exposure.

3. Final Answer:

   LoadBalancer Service -> Option A

4. Quick Check:

   Single app direct IP = LoadBalancer Service [OK]

Hint: Single app direct IP? Use LoadBalancer Service [OK]

Common Mistakes:

• Confusing Ingress with LoadBalancer for single app
• Choosing ClusterIP which is internal only
• Selecting NodePort which exposes on node ports, not external IP

2. Which of the following is the correct way to define a LoadBalancer service in a Kubernetes YAML manifest?

easy

A. apiVersion: networking.k8s.io/v1\nkind: LoadBalancer\nmetadata:\n name: my-service\nspec:\n selector:\n app: my-app\n ports:\n - protocol: TCP\n port: 80\n targetPort: 8080

B. apiVersion: v1\nkind: Service\nmetadata:\n name: my-service\nspec:\n type: Ingress\n selector:\n app: my-app\n ports:\n - protocol: TCP\n port: 80\n targetPort: 8080

C. apiVersion: v1\nkind: Service\nmetadata:\n name: my-service\nspec:\n type: NodePort\n selector:\n app: my-app\n ports:\n - protocol: TCP\n port: 80\n targetPort: 8080

D. apiVersion: v1\nkind: Service\nmetadata:\n name: my-service\nspec:\n type: LoadBalancer\n selector:\n app: my-app\n ports:\n - protocol: TCP\n port: 80\n targetPort: 8080

Solution

1. Step 1: Check apiVersion and kind

   Service resources use apiVersion: v1 and kind: Service.

2. Step 2: Confirm service type

   LoadBalancer type is set under spec.type to expose external IP.

3. Final Answer:

   YAML with kind Service and type LoadBalancer -> Option D

4. Quick Check:

   LoadBalancer service YAML uses kind Service and type LoadBalancer [OK]

Hint: LoadBalancer service uses kind Service and type LoadBalancer [OK]

Common Mistakes:

• Using kind LoadBalancer instead of Service
• Setting type to Ingress which is invalid for Service
• Confusing NodePort with LoadBalancer

3. Given this scenario: You have two apps, app1 and app2, and want to expose both externally using one IP with different paths. Which Kubernetes resource will handle this correctly?

medium

A. Two LoadBalancer Services, one for each app

B. One Ingress resource with path rules for app1 and app2

C. Two NodePort Services with different ports

D. One ClusterIP Service with multiple selectors

Solution

1. Step 1: Understand routing multiple apps on one IP

   Ingress allows routing multiple apps through one IP using path or hostname rules.

2. Step 2: Compare with LoadBalancer and others

   LoadBalancer exposes one app per IP; NodePort uses node ports, not one IP; ClusterIP is internal only.

3. Final Answer:

   One Ingress resource with path rules -> Option B

4. Quick Check:

   Multiple apps, one IP = Ingress with path rules [OK]

Hint: Multiple apps, one IP? Use Ingress with paths [OK]

Common Mistakes:

• Using multiple LoadBalancers wasting IPs
• Choosing NodePort which needs manual port management
• Thinking ClusterIP exposes externally

4. You created a LoadBalancer service but it never gets an external IP assigned. What is the most likely cause?

medium

A. You forgot to set spec.type to ClusterIP

B. Ingress controller is missing

C. Your Kubernetes cluster is running on a platform without LoadBalancer support

D. You used NodePort instead of LoadBalancer

Solution

1. Step 1: Check platform support for LoadBalancer

   Some clusters (like bare metal) do not support automatic external IP assignment for LoadBalancer services.

2. Step 2: Verify service type and controller

   Spec.type must be LoadBalancer, not ClusterIP; Ingress controller is unrelated; NodePort is different service type.

3. Final Answer:

   Platform lacks LoadBalancer support -> Option C

4. Quick Check:

   No external IP = platform lacks LoadBalancer support [OK]

Hint: No external IP? Check platform LoadBalancer support [OK]

Common Mistakes:

• Confusing ClusterIP with LoadBalancer type
• Assuming Ingress controller assigns LoadBalancer IP
• Mixing NodePort with LoadBalancer service

5. You want to minimize cloud costs while exposing 5 different apps externally. Which approach is best?

hard

A. Use one Ingress resource to route all 5 apps through a single LoadBalancer IP

B. Use ClusterIP services and expose each with a separate external IP

C. Use 5 NodePort services and expose nodes directly

D. Create 5 LoadBalancer services, one per app

Solution

1. Step 1: Understand cost implications of LoadBalancer IPs

   Each LoadBalancer service usually gets a separate external IP, increasing cloud costs.

2. Step 2: Use Ingress for cost efficiency

   Ingress lets multiple apps share one external IP with routing rules, saving costs.

3. Final Answer:

   One Ingress resource with single LoadBalancer IP -> Option A

4. Quick Check:

   Multiple apps, minimize cost = Ingress with one LoadBalancer IP [OK]

Hint: Multiple apps, one IP saves cost: use Ingress [OK]

Common Mistakes:

• Creating multiple LoadBalancers wasting IPs and money
• Using NodePort exposes nodes, less secure and complex
• ClusterIP cannot expose apps externally