← sources

Meme from Russian DevOps channel (autrotranslated to English) AKA "List of DevOps interview questions for October 2026"

source: reddit (reddit:devops)published 2026-10-12 19:39ingested 2026-10-13 02:37

retained text

* Do you have experience literally extinguishing a production environment?
* What is the difference between Disaster Recovery and Damage Assessment?
* What is the RTO for a data center after a direct drone hit?
* Can you debug infrastructure using satellite imagery?
* What would you do if your Kubernetes cluster suddenly became "bare metal" in the literal sense?
* How do you configure HPA so that pods can outrun drones?
* What would you do if not just a node, but the entire data center went down?
* Explain the difference between a Blue-Green Deployment and a "Red" air raid alert.
* How do you interpret the term "hot standby" if the backup data center is still on fire?
* What Prometheus alert would you set up for smoke coming from the server room?
* Have you ever migrated infrastructure using `kubectl evacuate`?
* What do you do if `terraform destroy` was executed without Terraform?
* Can a drone strike on the server room be considered a successful Chaos Engineering experiment?
* How many nines of availability remain for the service after losing two data centers?
* How do you distinguish a DDoS attack from an attack that renders a DDoS attack unnecessary?
* What do you do if your disaster recovery plan begins with the words "find a surviving server"?
* How do you explain to management that an Availability Zone is no longer available?
* Which metric do you choose for your SLO: uptime, MTTR, or the number of surviving racks?
* How do you ensure geo-redundancy when geography itself has become the primary risk factor?
* What do you do if you have an Active-Active setup, but both sites have switched to an Inactive-Inactive state?
* How do you configure fencing in a cluster when the fencing has already occurred physically?
* What do you do if your Witness in the third data center is the first to stop showing signs of life?
* Can a storage system be considered fault-tolerant if the data outlived the hardware, but the hardware didn't survive its operational lifespan?
* How do you calculate the "blast radius" when it’s no longer just a metaphor from the documentation?
* What is the point of synchronous replication if both data centers cease to exist synchronously?
* How do you implement automatic failover when the decision to switch over is made by the Ministry of Emergency Situations?
* Discuss what status should you set in Grafana: CRITICAL, DOWN, and NO LONGER EXISTS?
* How do you convince the CFO to approve a third data center when the first two have already been written off as CAPEX?
* What backup retention period do you recommend when the main issue is the lifespan of the storage hardware itself?
* How do you achieve "five nines" of availability when the only nines in the project are the digits in the recovery cost?

document fetched 2026-10-13 00:26 · canonical https://reddit.com/r/devops/comments/1x4bpht/meme_from_russian_devops_channel_autrotranslated