Low utilisation with terrible tail latency is not a contradiction. It is the CFS quota doing exactly what it was configured to do. Most candidates have never read nr_throttled.
an account raises the per-topic limit · no card
← Incident Response & Production Debugging / 12
hard★ EssentialNewGoogleUberDatadog
p99 spikes, probes time out, CPU utilisation reads 30 percent and nothing is OOMKilled. cpu.stat shows nr_throttled climbing. Explain it.
Low utilisation with terrible tail latency is not a contradiction. It is the CFS quota doing exactly what it was configured to do. Most candidates have never read nr_throttled.
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.
UP NEXT ON YOUR JOURNEY
Next in this trackCheckout is failing with 'connection pool exhausted', but the database CPU sits at 10 percent. Where is the problem?Next in this trackA busy node drops new connections intermittently; dmesg shows 'nf_conntrack: table full'. Walk me through what is happening.Next in this trackOne dependency slowed down and three services fell over completely. Walk me through the mechanics, then how you stop the next one.
DISCUSSION · 0
Nothing here yet. Say how you would answer it.