The DevOps & SRE Interview Roadmap: Infrastructure at Scale

The Shift from SysAdmin to SRE

The days of the traditional Systems Administrator manually configuring servers via SSH are over. Today, Infrastructure is Code.

Whether you are applying for a DevOps Engineer role or a Site Reliability Engineer (SRE) role, the SRE interview preparation process requires a deep understanding of cloud architecture, automation, and incident management.

1. Linux & Networking Fundamentals

Before you can orchestrate containers, you must understand the host OS and the network.

• Linux Internals: Expect questions on file permissions, processes (top, htop, strace), and system resources (CPU load vs Memory).
• Networking: You must deeply understand the OSI model. Be prepared to explain DNS resolution step-by-step, TCP handshakes, subnetting, and HTTP/HTTPS protocols.

2. Infrastructure as Code (IaC) & CI/CD

If you are clicking around the AWS console to provision servers, you are doing it wrong. Interviewers want to see that you can automate everything.

Common DevOps Interview Questions:

• "How do you manage state in Terraform?"
• "Explain the difference between a mutable and immutable infrastructure."
• "Design a CI/CD pipeline for a microservices architecture using GitHub Actions and Jenkins."

3. The Kubernetes Interview

Kubernetes (K8s) is the undisputed king of container orchestration. A modern kubernetes interview will test your practical knowledge of deploying and scaling applications.

Key Concepts:

• Pods vs Deployments vs StatefulSets: Know when to use which.
• Services & Ingress: How does traffic actually reach your Pod from the outside world?
• ConfigMaps & Secrets: How do you securely manage environment variables?
• Troubleshooting: "A Pod is stuck in CrashLoopBackOff. Walk me through exactly how you would debug it."

4. Reliability & Incident Response

This is the core of the SRE role, pioneered by Google. You must understand the difference between:

• SLI (Service Level Indicator): A direct measurement of behavior (e.g., HTTP 500 error rate).
• SLO (Service Level Objective): Your target goal (e.g., 99.9% of requests will succeed).
• SLA (Service Level Agreement): The legal/business contract if the SLO is not met.

Interviewers will also test your incident response. They might give you a scenario: "The primary database just went down during Black Friday. What do you do?" They are looking for a calm, methodical approach: mitigating the immediate impact, identifying the root cause, communicating with stakeholders, and writing a blameless post-mortem.

How to Practice

Theoretical knowledge is not enough for DevOps. You must build. Set up a mini K8s cluster locally using Minikube or Docker Desktop, write Terraform scripts to provision AWS resources, and practice explaining your architecture choices out loud using an AI Mock Interview platform.