DevOps and SRE troubleshooting guides
Field guides for broken infrastructure, and opinions on hiring the people who fix it.
Newest 20 of 48 posts. Search covers all of them.
Posts
From developer to SRE: the skill bridge that matters
How to become an SRE from a developer background: the skills that actually transfer, the gaps to close, and how to get incident reps without the title.
Kubernetes NetworkPolicy debugging: when pods can't talk
Kubernetes NetworkPolicy not working - or working too well? A test method for silent packet drops: default-deny surprises, selector typos, DNS egress.
Zombie processes: what defunct really means
Zombie processes explained: what <defunct> means, why kill -9 does nothing, how to find the negligent parent, and when a zombie process on Linux matters.
Quiz-based DevOps screening: where MCQs work, where they lie
When is a TestGorilla alternative worth it? Where MCQ screening genuinely works, where it lies about senior DevOps skill, and the escalation ladder.
Linux interview questions: scenarios from real 3 AM pages
Linux interview questions drawn from real 3 AM pages: disk full, runaway load, OOM kills. What strong answers sound like and how to grade the weak ones.
PVC stuck in Pending and volume mount failures
A PVC stuck in Pending in Kubernetes: StorageClass typos, missing provisioners, WaitForFirstConsumer, access modes, and the mount failures that follow.
What hiring can learn from incident reviews
Your engineering hiring process could learn a lot from your incident reviews: blameless analysis, process evidence, and a bias for how over whether.
Requests and limits: the misconfigurations that page you
Kubernetes resource limits and requests, minus the folklore: CPU throttling, limit=request tradeoffs, missing requests, and the LimitRange safety net.
Linux DNS not resolving: resolv.conf and systemd-resolved
Linux DNS not resolving? Trace the real resolution path - nsswitch, /etc/hosts, resolv.conf, systemd-resolved's stub - and why dig works but curl fails.
CoderPad vs Faultybox: pair programming vs incident response
CoderPad vs Faultybox, honestly: a superb pairing editor and a CoderPad alternative for incident response - different jobs, and when you need each one.
How to run a live debugging interview that isn't torture
A practical guide to running a live debugging interview: environment prep, hint policy, think-aloud prompts, timeboxing, and a scoring sheet that works.
Ingress 502, 503, 504: which layer is lying to you
Kubernetes ingress 502, 503, and 504 errors decoded: which layer failed, which logs to read, and the fastest path from load balancer to pod to find it.
Proctoring is an arms race you're funding
Interview proctoring pits detection against generation, and detection always trails. Every false positive burns a candidate you wanted. Stop funding it.
How to practice Linux troubleshooting deliberately
Practice Linux troubleshooting the deliberate way: break-it-yourself drills, a skill ladder from beginner to on-call ready, and feedback loops that stick.
Evicted pods: node pressure and who gets killed first
Kubernetes pod evicted? How node-pressure eviction works, what the eviction message means, why QoS class decides who dies first, and how to clean up.
Address already in use: who owns this port?
bind: address already in use - find the process holding the port with ss and lsof, tell TIME_WAIT from a live listener, and fix it without rebooting.
Codility alternatives for infrastructure hiring
Codility alternatives for infrastructure hiring: why algorithmic screens miss ops skill, and how to test DevOps and SRE candidates on real systems.
Platform engineer interview questions that test real work
Platform engineer interview questions should test the platform, not memory. Sample exercises for IaC review, Kubernetes debugging, and API judgment.
Debugging Kubernetes DNS: from nslookup to CoreDNS
Kubernetes DNS troubleshooting from symptom to CoreDNS: reproduce with a debug pod, decode the ndots trap, read CoreDNS logs, check upstream resolvers.
Why we log commands, not screen video
Interview recording doesn't need a camera. A timestamped command log is searchable, costs kilobytes, and can't capture anything outside the lab.