AWS scenario interview questions

AWS questions for DevOps and SRE roles tend to start with a symptom: a bill that jumped, a database that ran out of connections, a bucket someone made public. The interviewer wants to see you find the cause and weigh cost, security and blast radius.

Each question below is the exact prompt from a DevOps Ledger challenge. Decide what you would do, then open the challenge to defend it.

13 scenarios: 3 beginner, 6 intermediate, 4 advanced. Answer keys reviewed by Branko Petric.

What interviewers probe

  • Finding what drives a cost spike before cutting anything
  • IAM roles and short-lived credentials instead of long-lived keys
  • Connection limits, pooling and retries against managed databases
  • At-least-once delivery and idempotent consumers

The questions

  1. 01beginner9 min

    Your S3 Website Returns 403. Make the Bucket Public?

    Your app gets 403 AccessDenied fetching objects from a private S3 bucket. A teammate says just turn off Block Public Access and make the bucket public so it works. Do you agree, and what do you do instead?

    Practice this scenario
  2. 02beginner9 min

    The t3 Instance Got Slow. Just Size It Up?

    Your service runs fine for a while, then gets slow under sustained traffic on a t3.medium. CPU looks pinned near a flat ceiling. A teammate says just upgrade to a bigger t3. What is really happening, and what do you do?

    Practice this scenario
  3. 03beginner10 min

    Access Keys Are Hardcoded on the EC2 Box. Just Rotate Them?

    Your app on EC2 talks to AWS using long-lived IAM user access keys stored in a config file on the instance, and one set just leaked. A teammate says generate new keys and paste them in. Do you agree, and what should you do instead?

    Practice this scenario
  4. 04intermediate12 min

    Lambda Is Exhausting Your RDS Connections. Raise the Limit?

    During traffic spikes your Lambda functions start failing with too many connections errors against RDS. A teammate says just raise max_connections on the database. Do you agree, and what do you do instead?

    Practice this scenario
  5. 05intermediate12 min

    The NAT Gateway Bill Exploded. Add More Gateways?

    Cost Explorer shows NAT Gateway data processing charges have become your largest line item. Private-subnet instances pull large images and data from S3 and ECR. A teammate says it is just bandwidth, add another NAT gateway. What do you do?

    Practice this scenario
  6. 06intermediate13 min

    SQS Is Delivering the Same Message Twice. Is the Queue Broken?

    Your workers sometimes process the same SQS message twice, leading to duplicate charges and emails. A teammate says standard SQS is broken and wants to move everything to FIFO. What is actually happening, and what do you do?

    Practice this scenario
  7. 07advanced15 min

    DynamoDB Is Throttling but Capacity Looks Idle. Crank It Up?

    DynamoDB is returning throttling errors on certain requests, yet the table-level consumed capacity is far below what is provisioned. A teammate says just raise the capacity or switch to on-demand. What is really happening, and what do you do?

    Practice this scenario
  8. 08advanced15 min

    Cross-Account S3 Access Still Fails. Just Open It Up?

    You granted a partner account access to your S3 bucket with a bucket policy, but their role still gets AccessDenied reading the objects, which are encrypted with a customer-managed KMS key. A teammate says just make the bucket public and let the key allow everyone. What do you do?

    Practice this scenario
  9. 09intermediate13 min

    Finance Wants Everything on Spot Instances. Including the Database.

    Finance has mandated a cloud cost cut and points at spot instances: 70 percent cheaper, and 'the cloud provider says interruptions are rare'. They want everything moved, including the primary Postgres server. Do you push back, and where do you draw the line? Defend your call.

    Practice this scenario
  10. 10advanced15 min

    Your Primary Region Is Down. Fail Over Now?

    Your primary cloud region is having a major outage and your app is degraded. You have a standby in another region with an async database replica. A teammate says just flip DNS to the standby right now. Do you fail over, and how?

    Practice this scenario
  11. 11advanced15 min

    An EKS Pod Needs AWS Access. Attach It to the Node Role?

    A pod in your EKS cluster needs to read from S3. A teammate says just add the S3 permission to the node instance role so the pod can use it immediately. Do you agree, and what should you do instead?

    Practice this scenario
  12. 12intermediate12 min

    Deploy From CI: Paste AWS Keys Into GitHub Secrets?

    The deploy job needs AWS permissions. A teammate created an IAM user, generated an access key and secret, and wants to store them as GitHub Actions secrets so the pipeline can deploy. Do you agree, and what do you do instead?

    Practice this scenario
  13. 13intermediate12 min

    Every Rolling Deploy Causes 502s. Add Ingress Retries?

    Every rolling deploy of your API causes a short burst of 502s at the load balancer, then it settles. A teammate wants to turn on retries at the ingress and move deploys to the night window. Do you agree, and what do you do instead?

    Practice this scenario

Worked answers

Free AWS guides

Practice defending your decisions

DevOps Ledger is a scenario-based DevOps and SRE interview practice platform: you pick a real production incident, defend your decision in a chat against a senior-engineer AI that pushes back, and get scored against an answer key written before you started.

A general chatbot makes up an answer on the spot. DevOps Ledger grades you against an answer key written before you started.

One payment. 12 months of access. No subscription, nothing renews.

Get full access - $19

Other topics