DevOpsInterviewPrep logo
← Containers & Kubernetes / 46
easyNewDatadogUberNVIDIA

Worker nodes slowly fill their disks until DiskPressure appears. How does the kubelet decide what to clean up?

Two cleanup mechanisms hide behind one taint: image garbage collection removes unreferenced images by LRU at threshold crossings, while eviction kills pods. Candidates who conflate them fail this question quietly.

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.

Two cleanup mechanisms hide behind one taint: image garbage collection removes unreferenced images by LRU at threshold crossings, while eviction kills pods. Candidates who conflate them fail this question quietly.

20 answers per topic instead of 10, and your progress kept · no cardor unlock all 390 remaining answers · ₹2,000 / $25
UP NEXT ON YOUR JOURNEY
Next in this trackWith tens of thousands of services, DNS lookups start timing out for five seconds at a stretch. Diagnose and fix Kubernetes DNS at scale.Next in this trackThe HPA shows a CPU number. Trace it from the container to the scaling decision, and tell me where that number lies.Next in this trackSidecar proxies cost us real memory per pod. Do ambient and Cilium-style meshes actually remove that cost, and what do they trade?
DISCUSSION · 0

Nothing here yet. Say how you would answer it.