must-triage

When things go wrong with OpenShift or OpenShift Container Storage, we sometimes use must-gather or ocs-must-gather to collect logfiles and other documents to diagnose issues. While those tools are useful in that they can show us a lot about the state of the system(s), they don't offer anything to help process that information.

must-triage is a very small utility to aid in identifying problems recorded in must-gather output.

Features

Currently, must-triage knows about the following potential issues:

OpenShift (OCP)

• Unparseable YAML files
• Pods not in ready state
• Operators who haven't reported success
• YAML files which contain a top-level error field

OpenShift Container Storage (OCS)

• Empty JSON files
• Unparseable JSON files
• Ceph health not HEALTH_OK
• Logfiles which mention a panic

Teaching it to identify new issues is not difficult; PRs warmly welcomed!

Installation

pip install must-triage

Usage

must-triage /path/to/must-gather/output