🌐 Networking for DevOps – Interview Scenario Q&A (In-Depth)

---

1️⃣ Scenario: Website is Not Reachable From Browser

Question:
Your application is running on a server, but users cannot access it from the browser. How will you troubleshoot?

✅ Step-by-Step Answer (Interviewer Loves This)

1. Check DNS Resolution

nslookup example.com
dig example.com

• If DNS fails → Domain not mapped correctly

2. Check Network Connectivity

ping example.com
traceroute example.com

• Identifies where the packet is dropping

3. Check Port Accessibility

telnet example.com 80
nc -zv example.com 443

4. Check Firewall / Security Groups

• Linux firewall:

sudo iptables -L
sudo ufw status

• Cloud SG/NACL rules

5. Check Application Binding

ss -tulnp | grep 80

• App must bind to 0.0.0.0, not 127.0.0.1

6. Check Load Balancer (if any)

• Health checks

• Target group status

👉 Conclusion:
Networking issue could be DNS, firewall, port binding, or load balancer misconfiguration.

---

2️⃣ Scenario: Application Works Locally but Fails in Production

Question:
Your app works on localhost but fails after deployment. Why?

✅ Possible Reasons

🔍 Debug Commands

curl localhost:8080
curl server-ip:8080

---

3️⃣ Scenario: Kubernetes Pod Cannot Reach External Internet

Question:
A pod inside Kubernetes cannot access external APIs. What could be wrong?

✅ Root Causes

1. No NAT Gateway

   • Pods in private subnet need NAT for outbound internet

2. Network Policies Blocking Traffic

kubectl get networkpolicy

3. DNS Issue

kubectl exec pod -- nslookup google.com

4. CoreDNS Not Working

kubectl get pods -n kube-system

5. Security Group Rules

• Node SG must allow outbound traffic

👉 DevOps Insight:
Kubernetes networking depends heavily on cloud VPC + CNI plugin.

---

4️⃣ Scenario: High Latency in Application

Question:
Users complain about slow response times. How do you troubleshoot networking?

✅ Investigation Flow

1. Check Latency

ping server

2. Check Packet Loss

mtr example.com

3. Check Load Balancer

• Uneven traffic distribution

4. Check MTU Mismatch

• Common in VPN / container networking

5. Check TLS Handshake Time

curl -w "@curl-format.txt" -o /dev/null -s https://example.com

---

5️⃣ Scenario: SSH Connection Refused

Question:
You cannot SSH into a server. What will you check?

✅ Checklist

1. Correct Port

ssh user@ip -p 22

2. Firewall Rules

iptables -L

3. SSH Service Status

systemctl status sshd

4. Security Group

• Port 22 allowed from your IP?

5. Fail2ban Blocking IP

sudo fail2ban-client status

---

6️⃣ Scenario: Load Balancer Shows Targets as Unhealthy

Question:
ALB/NLB shows unhealthy instances but app is running.

✅ Common Reasons

• Health check path incorrect (/health)

• App returns 403 or 500

• Wrong port mapping

• Firewall blocks LB IP range

👉 Fix

curl http://instance-ip:port/health

---

7️⃣ Scenario: CI/CD Pipeline Cannot Access Server

Question:
Your GitHub Actions/Jenkins cannot deploy to server.

✅ Debug Steps

• Whitelist CI IP range

• Check outbound rules

• VPN or Bastion Host required?

• SSH key mismatch

---

8️⃣ Scenario: Microservices Cannot Talk to Each Other

Question:
Service A cannot reach Service B.

✅ Possible Causes

• Wrong service name (DNS)

• Network policy blocking

• Wrong port exposure

• mTLS misconfiguration

curl http://service-b.namespace.svc.cluster.local

---

9️⃣ Scenario: Public IP Changed After Restart

Question:
Your server IP changes after restart. Why?

✅ Explanation

• Using dynamic public IP

• Fix by:

  • Elastic IP (cloud)

  • Load balancer

  • DNS instead of IP

---

🔟 Scenario: Zero-Downtime Deployment Networking Role

Question:
How networking helps zero-downtime deployment?

✅ Concepts Used

• Load balancer health checks

• Blue-Green deployment

• Canary routing

• DNS TTL control

---

🔥 MUST-KNOW COMMANDS (Interview Rapid Fire)

ip a
ip route
ss -tulnp
netstat -rn
tcpdump -i eth0
curl -I
traceroute
nslookup
dig

---

🧠 Interviewer Bonus Tip

When answering:

❌ “It’s a networking issue”
✅ “I’ll verify DNS → connectivity → port → firewall → application binding”

This structured thinking is what interviewers want from DevOps engineers.