Meta DevOps / SRE Interview Questions

1.How do you prioritize your work when you have multiple production issues, feature requests, and infrastructure improvements competing for your time?

2.Walk me through a time you had to quickly diagnose and fix a production outage. What was your process and how did you improve your response afterward?

3.Tell me about the last time you had to implement a critical infrastructure change under tight time pressure. How did you balance speed with reliability?

4.Tell me about a time you had to give tough feedback to a teammate about their on-call performance or operational practices. How did you approach it?

5.Describe a time you pushed for adopting a new technology or tool that your team was hesitant about. What was your approach and what happened?

6.Describe a time when an infrastructure decision you made had broader implications for user experience or privacy. How did you think through those considerations?

7.Write a script to automatically detect and restart unhealthy containers in a Kubernetes cluster based on custom health metrics beyond the default health checks.

8.Design a log aggregation system that can handle logs from all of Facebook's web servers during peak traffic while allowing engineers to query recent logs in real-time.

9.We're rolling out a new version of HHVM across thousands of web servers that serve Facebook's news feed. How would you design the deployment strategy to minimize user impact while catching issues early?

10.You need to implement monitoring for WhatsApp's message delivery system to track end-to-end latency across different regions. What metrics would you collect and how would you structure the monitoring?

11.Instagram's image upload service is experiencing intermittent timeouts during peak hours. The service handles millions of uploads daily and integrates with multiple backend systems. Walk me through your debugging approach.

12.TAO, our social graph store, is seeing increased read latency during peak traffic. You have read replicas, but queries are still slow. How would you approach optimizing this?

13.Design a configuration management system that can push updates to all of Facebook's web servers within 2 minutes during an outage. How do you ensure consistency when some servers might be unreachable?

14.Design a system to automatically detect and mitigate DDoS attacks targeting Threads while preserving the real-time nature of text conversations and trending topics.

15.WhatsApp needs to migrate its message storage from one data center to another without any downtime. Design the migration strategy for petabytes of message data while maintaining end-to-end encryption.

16.Instagram's photo processing pipeline is backing up during major events like New Year's Eve when upload volume spikes 50x. How would you design auto-scaling that maintains image quality while handling the burst?

17.Design a chaos engineering system for Meta's infrastructure that can safely test failure scenarios across Facebook, Instagram, and WhatsApp without impacting real users.

18.Meta Quest headsets need to download game updates and social features while maintaining low latency for VR experiences. Design a content delivery system that prioritizes interactive VR traffic over background downloads.

19.Tell me about a time when you had to convince multiple engineering teams to adopt a new operational practice or tool. How did you get buy-in without having direct authority over them?

20.Tell me about a time you had to lead an on-call rotation or operational practice change when team members had different levels of experience. How did you ensure everyone could be effective?

21.Describe a situation where you had to make a quick decision during an outage that affected user experience. How did you balance moving fast with getting the right outcome?

22.Tell me about a time you had to push back on a product team's feature request because of infrastructure constraints. How did you handle that conversation?

23.Describe a time when you took ownership of a cross-team infrastructure problem that wasn't clearly your responsibility. What motivated you to step up?

24.Describe a time when you had to challenge a senior engineer or architect's technical decision that you believed would create operational problems. How did you approach that conversation?

25.Facebook's news feed ranking algorithm needs to process 100 billion posts daily to generate personalized feeds for 3 billion users. Estimate how much compute capacity we'd need if we switched from batch processing to real-time ranking.

26.Meta's data centers use 5 TWh of electricity annually. If we could reduce server CPU utilization by 3% through better load balancing, estimate the cost savings and environmental impact. How would you measure and validate these improvements?

27.Instagram Reels engagement dropped 8% globally overnight with no code deployments or known incidents. Walk me through how you'd investigate this and what you'd check first.

28.Threads is launching a new real-time trending topics feature that could generate 10x more database writes during viral events. Estimate the infrastructure cost increase and how you'd justify this investment to leadership.

29.You're designing capacity planning for Meta Quest's multiplayer VR services. During holiday launches, concurrent users spike 20x normal levels for 2-3 hours. How do you size infrastructure to handle this cost-effectively?

30.WhatsApp processes 100 billion messages daily. If we wanted to add message search functionality, estimate the storage and compute costs, and explain how you'd architect this without impacting message delivery performance.