Securing an internship in DevOps in 2026 wasn't something I had perfectly planned. In fact, if I had to describe it honestly, it felt like one of those moments where people say, "You have to be at the right place at the right time." And somehow, that's exactly what happened to me.

The opportunity came unexpectedly on January 8th 2026. An email popped up in our college mail inbox with the subject line: "Urgent: DevOps Hiring - Interviews Tomorrow." It was for DevOps SME role at Scaler (formerly InterviewBit) ** - 6 Month Internship** with PPO on the line.

Before the first round, there was an initial filtering process conducted by our placement cell. I'm not sure how many students applied in total, but after the resume screening, only 11 students were shortlisted for the 1st Round.

The First Round: The one that I thought I fucked up!

The first round of the interview was conducted by a third-party company called Interview Vector, which Scaler had outsourced the initial screening to. At that time, I didn't know much about the interviewer himself. Later, I discovered that he was a Member of Technical Staff at Oracle. Earlier in his career, he had worked as a DevOps engineer at Flipkart, and before transitioning fully into DevOps, he had experience as a network engineer and a Linux system administrator.

The interview lasted 1 hour and 15 minutes.

Here is how it went:

First I was given a shell scripting question which was to be solved to move ahead in the interview:

Shell Scripting Ques.
 
Write a script to count how many times each word appears in a file (notes.txt), ignoring case sensitivity.
 
Given Input in File:
Hello
HELLO
World
Devops
world
hellO
 
Expected Output Format:
hello : 3
world : 2
devops : 1

Technical Questions Asked in the First Round

After the shell scripting problem, the interviewer moved into a series of deep technical questions. These questions covered multiple domains, including Networking, Operating Systems, Linux, Docker, Cloud (AWS), and Terraform.

Below is the complete list of questions I was asked during the interview:

DNS and Networking Fundamentals

• What is DNS? Explain DNS internals.

• How does DNS caching work?

• What is TTL (Time To Live) in DNS?

• Explain browser DNS cache and OS DNS cache.

• What is Reverse DNS?

• What is Wrong DNS?

TCP/IP and Networking Concepts

• How does TCP work? What are the common use cases of TCP?

• Explain the TCP three-way handshake.

• What are the main TCP flags?

• What is the difference between TCP and UDP?

Operating Systems and Resource Management

• What is a deadlock? What are the four necessary conditions for a deadlock?

• What is CPU throttling? Why does it happen?

• What is memory throttling?

• What is the difference between CPU throttling and memory throttling? - Long discussion went on it!

• Why does your browser lag when you open 40–50 tabs or websites at once?

Linux and Process Monitoring

• How do you identify processes consuming high CPU usage?

• After identifying a high CPU usage process, how would you monitor it? Which linux cmds you will use?

• What is OOM (Out Of Memory)?

Docker and Container Monitoring

• How do you monitor Docker container logs?

• How would you monitor logs if there are 20 containers running simultaneously?

Cloud and AWS Scenario-Based Questions

• The interviewer framed a real-world scenario and indirectly asked me to explain the architecture of VPC Peering.

• How would you manage log files in AWS? This question was scenario-based and indirectly focused on using S3 lifecycle policies to transition logs across different storage classes such as:

  • Standard Storage

  • Glacier Instant Retrieval

  • Glacier Deep Archive

EC2 and Cloud Cost Optimization

• What are the different types of EC2 instances?

• The interviewer gave a scenario-based question to evaluate my understanding of the use cases and differences between On-Demand instances and Spot instances.

Infrastructure as Code (Terraform)

• Explain Terraform state management.

Interview Difficulty and Key Takeaway

This round was extremely intense. I had never experienced an interview of this length before - it lasted around one and a half hours. By the end of it, I felt mentally drained. However, despite how exhausting it was, it turned out to be an incredible learning experience.

After the first interview, out of eleven, 5 students were selected for the second round.

The Second Round: The Best One!

The second round started with a brief introduction about myself. The interviewer asked about my background, my past summer internship at Garaaz, and my work with MongoDB Atlas Charts. We discussed the kind of work I had done, the problems I had worked on, and the tools I had used.

Since I had mentioned product management experience in my resume, the interviewer also asked a few questions about that. It was a very brief discussion focused on understanding my responsibilities and how I approached problem-solving in that role.

After this initial discussion, the technical part of the interview began.

Multi-Format Technical Interview

The second round was different from the first. Instead of focusing on a single type of question, the interviewer evaluated me through multiple formats, including practical tasks, live modifications, rapid-fire commands, system design, and hands-on terminal exercises.

The round started with a Live Dockerfile discussion. The interviewer shared his screen, asked me to read the Dockerfile opened in his IDE, and explain what it was doing. He then made changes in real time and asked me to explain the impact of those changes on the build process and container behavior.

Next, I was asked to write a short Terraform HCL configuration to create an S3 bucket in the Mumbai region (ap-south-1). The interviewer expected a short and Terraform code to define the provider and create the S3 bucket resource.

After that, there was a Rapid-Fire Linux Round where I was asked multiple networking-related commands. I had to write and execute them directly on the terminal provided.

The interview then shifted to testing my Cloud Architecture understanding. This was an discussion where I was asked how would I design a 3-tier application architecture within a single VPC, including the Web tier, Application tier, and Database tier. I also had to explain how to configure components such as:

• NAT Gateway

• Internet Gateway

• Security Groups / Firewalls

The discussion focused on secure communication, network isolation, and proper routing.

Finally, I was given a hands-on Log Processing task on the terminal. I was provided with a log file and asked to:

• Extract only GET API requests

• Extract the IP addresses from those requests

• Remove duplicate IP addresses

• Output only the unique IP addresses into a final result file

Technical Questions Asked in the Second Round

Below is the complete list of questions I was asked during the interview:

CI/CD and DevSecOps

• Explain the complete end-to-end CI/CD pipeline, including DevSecOps tools, Docker, GitLab CI, and deployment to EC2.

• Questions related to security tools such as Trivy, SonarQube, and SAST were discussed as part of the pipeline.

Containerization (Docker)

• What is the difference between:

  • Docker Image

  • Dockerfile

  • Docker Container

• What is a multi-stage Docker build?

• What is the purpose of a .dockerignore file?

• What is the difference between Amazon ECR and GitLab Container Registry?

Kubernetes Architecture and Concepts

• When a user runs a Kubernetes command (for example, kubectl get pods), what happens internally?

  • This question indirectly tested my understanding of the full Kubernetes architecture and workflow.

• What is a ReplicaSet?

• What is a Deployment?

• Explain the following Kubernetes Service types:

  • ClusterIP

  • NodePort

  • LoadBalancer

• Which Kubernetes component resembles the state file concept in Terraform? -- etcd

Terraform and State Management

• How do you manage the Terraform state file?

  • This was discussed through a scenario involving remote backends such as S3 and state locking mechanisms.

• Follow-up discussion indirectly testing knowledge of Terraform drift.

• How do you recover from a corrupted Terraform state file?

Deployment Strategies

• What is Blue-Green deployment?

• What is Canary deployment?

Networking and OS Fundamentals

• Explain all seven layers of the OSI model and their purposes.

• What is a VPC?

Using the CIDR block 10.0.0.0/16, I was asked to explain:

• CIDR notation

• Network bits and host bits

• Subnet masks

This round lasted around 1 hour. Before this interview, during my preparation, I had anticipated many of the topics that might be asked. I had even conducted mock interviews with my friends and discussed these questions with them.

Interestingly, around 80% of the questions I had expected were actually asked in the interview. Because of this preparation, I felt much more confident and comfortable throughout the round.

Out of all the rounds, this was my strongest performance. I was able to answer most of the questions clearly and handle the discussions well. This is how the second round went.

After the second interview, out of 5, 3 students were selected for the third round.

The Third and the Final Round!

The third and final round lasted around 20–25 minutes. It was taken by the DevOps Lead at Scaler.

He started by asking me to rate myself on the all DevOps skills I had listed on my resume - AWS, K8s, Docker, Terraform, Linux, Gitlab-CI, Shell Scripting.

Then, I was given a shell scripting question:

Shell Scripting Ques.
 
Write a shell script that accepts a directory as input and lists all log files older than 7 days for cleanup.

Technical Questions Asked in the Final Round

After the shell scripting problem, the interviewer asked a set of technical questions.

Below is the complete list of questions I was asked during the interview:

• What is Ansible, and have you worked with Ansible?

• What happens internally when a Docker container is created?

• How do you attach S3 access to an EC2 instance?

• How does EC2 securely access S3?

• Explain the role of IAM in EC2 to S3 access.

• How does EC2 access S3 using a VPC Endpoint?

• How do you manage Terraform state files using S3 and DynamoDB?

At the very end of the interview, after an entire discussion, he asked me something completely unexpected:

"If there is heavy rain and a football match gets cancelled, what will you do?"

For a second, I genuinely thought I had misheard the question. After 3 rounds multiple weeks of prep, grilling across rounds, this was the last thing I expected.

My first instinct was to say the obvious - football matches usually don't get cancelled because of rain. He smiled and laughed at that, but then asked me to assume that it did get cancelled and explain what I would do.

So I went along with the scenario and answered that I would focus on handling the situation properly - informing spectators, arranging refunds, trying to reschedule the match, and exploring alternatives like meet-and-greet sessions if possible.

It was a strange but memorable way to end the interview. After testing my technical skills, I think he just wanted to see how I think in unexpected situations when there's no predefined answer or documentation to rely on.

And with that, the final round ended.

The Final Result!

The interview process was spread across multiple dates:

• First Round - 10th January 2026

• Second Round - 13th January 2026

• Final Round - 21st January 2026

After all three rounds of getting grilled, second-guessing myself, and refreshing my inbox more times than I'd like to admit, the final result came out on 23rd Jan 2026.

Out of everyone who went through the process, only 1 person was selected - and this time, it was me.

I was the only candidate who cleared all three rounds. It felt unreal, especially because my 6th semester had just begun. While most of us were still settling into the semester, I had already gotten placed. I became one of the earliest students in my batch to receive an offer.

When the selection email was sent to everyone in the university, my phone suddenly started receiving messages. Friends, classmates, and seniors started congratulating me. It felt unreal, especially because just weeks before that, I was sitting in my room preparing, doubting myself, and hoping things would work out.

This was my first interview. First company. First real test of everything I had been learning.

And in the end, it all paid off.

I joined Scaler on-site in Bangalore on 28th January 2026.

Looking back, it all started with one unexpected email and turned into one of the most memorable moments of my life.