Pod State & Symptoms
Diagnostic Triage & Commands
Analyzing…
Remediation Steps
Recommended Diagnostic Commands
Understanding Operator Pod Failure Modes
Kubernetes operator pods differ from typical stateless web workloads. Operators run continuous reconciliation loops watching Custom Resource Definitions (CRDs). When an operator pod fails, it halts automation across all associated managed resources (databases, queues, or microservices).
Why Exit Code 137 Differs from Exit Code 1
Exit code 137 indicates the Linux kernel Out-Of-Memory (OOM) killer dispatched SIGKILL (signal 9 + 128 = 137) because the container process consumed more RAM than allowed in "resources.limits.memory". In contrast, exit code 1 indicates application-level crash (such as an unhandled Golang panic or invalid RBAC permission error while attempting to list CRDs).
Execution Safety Guarantee
All kubectl commands displayed by this tool are generated dynamically inside your browser. Operatorpod.net never connects to your cluster, never requests credentials, and never executes commands on your behalf.
A Systematic Diagnostic Playbook for Operator Pods
When an operator pod crashes, debugging requires tracing container lifecycle events, checking kernel termination signals, and reviewing custom resource controller logs. The Operator Pod Troubleshooter guides you through selecting observed phases, exit codes, and probe indicators, generating custom read-only kubectl triage commands tailored directly to your namespace and workload.
Differentiating Pod Failure Modes
- CrashLoopBackOff (Exit Code 1): The container process started, failed an internal initialization check (e.g. invalid CRD connection or missing RBAC permissions), and exited with an application-level error.
- OOMKilled (Exit Code 137): The container exceeded its memory limit or the host node experienced severe memory pressure, triggering Linux kernel SIGKILL.
- ImagePullBackOff: The kubelet was unable to pull the container image due to invalid repository names, missing imagePullSecrets, or registry rate limits.
- Evicted: The node triggered pod eviction due to DiskPressure, MemoryPressure, or PIDPressure.