DevOpsInterviewPrep logo
AI & GPU Infrastructure / 11
hard★ EssentialNewDatabricksNVIDIAGoogle

A vLLM pod serving an LLM OOMs at a request rate it handled fine yesterday, dropping in-flight requests. Walk me through it.

CPU memory intuition fails on inference servers. The scarce resource is KV-cache, requests arrive with unseen context lengths, and the fix is queue-depth scaling plus admission control, not bigger pods.

Updated Sep 2026 · Grounded in researched DevOps, SRE and platform engineering interview loops, written to a senior-engineer editorial bar, and never padded to hit a word count.

CPU memory intuition fails on inference servers. The scarce resource is KV-cache, requests arrive with unseen context lengths, and the fix is queue-depth scaling plus admission control, not bigger pods.

an account raises the per-topic limit · no card
UP NEXT ON YOUR JOURNEY
DISCUSSION · 0

Nothing here yet. Say how you would answer it.