HA AI Clusters - SRE Engineer - General questions
retained text
Hey guys, quick question for anyone in the AI or infrastructure space. I was never really interested in running AI models on my homelab cluster, but this job offer looked too good to pass up. I have a technical interview coming up soon for an SRE role on High Availability AI clusters, and honestly I have zero clue what to expect or what kind of questions they might throw at me. The position is completely remote, focusing on monitoring, troubleshooting, and keeping things running via SSH without any hands-on access to physical hardware. Core stack is mostly standard Linux, Docker, Kubernetes, and networking. If anyone here works with HA AI infrastructure or went through a similar interview process, what do they actually ask about? Do they stick to standard Linux and networking troubleshooting, or do they push hard on AI-specific tech like GPU drivers, Infiniband, NCCL, or CUDA failures? Any heads up or real-world examples would be awesome. Thanks in advance!
document fetched 2026-10-26 00:24 · canonical https://reddit.com/r/devops/comments/1xffaoh/ha_ai_clusters_sre_engineer_general_questions