Runbooks

RUNBOOKS

Troubleshoot with a plan.

Step-by-step operational runbooks for diagnosing common infrastructure failures.

KUBERNETES

Kubernetes Troubleshooting

Start from the symptom, collect evidence, then work toward the underlying cause.

Container keeps restartingCrashLoopBackOff →Image will not pullImagePullBackOff →Container runs out of memoryOOMKilled →Pod will not schedulePending Pods →
K8s

CrashLoopBackOff

Repeated container crashes: logs, exit codes, probes, configuration and dependencies.

Open runbook →IMG

ImagePullBackOff

Bad image names, registry authentication, networking and pull failures.

Open runbook →MEM

OOMKilled

Memory limits, usage patterns, runtime settings and application behavior.

Open runbook →SCH

Pending Pods

Scheduler events, resources, taints, affinity, PVCs and quotas.

Open runbook →SVC

Service Unreachable

Selectors, endpoints, ports, DNS, NetworkPolicies and application binding.

Open runbook →
LINUX

Linux Troubleshooting

Production-focused runbooks for host-level resource incidents.

MEM

High Memory / OOM Killed

Prove an OOM event, inspect memory pressure and identify runaway workloads.

Open runbook →DSK

Disk Usage / No Space Left

Move from filesystem pressure to the directory, file or process responsible.

Open runbook →CPU

High CPU Usage

Separate real CPU saturation from I/O wait, then identify the process or thread responsible.

Open runbook →SLOW

Server Running Slow

Triage CPU, memory, disk, I/O and network pressure before choosing a deeper runbook.

Open runbook →KILL

Process Keeps Getting Killed

Distinguish OOM kills, signals, crashes and service-manager restarts using evidence.

Open runbook →I/O

Disk I/O Troubleshooting

Interpret iowait, device latency, queueing and per-process I/O without relying on one metric.

Open runbook →PORT

Address Already in Use

Find the listener, understand its bind address and resolve port conflicts safely.

Open runbook →