Skip to content

Collecting Debug Information

This guide explains how to collect diagnostic information from your rhaii-on-xks deployment for troubleshooting or to share with Red Hat support.

Quick Start

./scripts/collect-debug-info.sh

Output is saved to /tmp/rhaii-on-xks-debug-<timestamp>/.

To specify a custom output directory:

./scripts/collect-debug-info.sh /path/to/output

What Is Collected

The script collects the following information from your cluster:

Step Component Data Collected
1 Cluster Kubernetes version, cluster info, node list
2 Helm All Helm releases across namespaces
3-4 cert-manager Operator, controller, and webhook pod status and logs
5 Certificates ClusterIssuers, Issuers, Certificates, CertificateRequests
6-7 Istio / Sail Operator Operator and istiod pod status and logs, Gateways, HTTPRoutes, InferencePools, GatewayClasses
8-9 LWS Operator Operator pod status and logs, LeaderWorkerSets, webhook configs
10-11 KServe Controller pod status and logs, LLMInferenceServices, LLMInferenceServiceConfig
12 Events Warning and error events from all operator namespaces

Namespaces Inspected

  • cert-manager-operator
  • cert-manager
  • istio-system
  • openshift-lws-operator
  • opendatahub

Output Files

rhaii-on-xks-debug-<timestamp>/
├── cluster-info.txt                  # Kubernetes version, nodes
├── helm-releases.txt                 # All Helm releases
├── cert-manager-status.txt           # Pods, deployments, CertManager CR
├── cert-manager-operator-logs.txt    # Operator logs
├── cert-manager-controller-logs.txt  # Controller logs
├── cert-manager-webhook-logs.txt     # Webhook logs
├── certificates-issuers.txt          # All certs, issuers, ClusterIssuers
├── istio-status.txt                  # Pods, Istio CR, Gateways, HTTPRoutes
├── istiod-logs.txt                   # istiod logs
├── sail-operator-logs.txt            # Sail operator logs
├── lws-operator-status.txt           # Pods, LeaderWorkerSets, webhooks
├── lws-operator-logs.txt             # LWS operator logs
├── kserve-status.txt                 # Pods, LLMInferenceServices, Gateway
├── kserve-controller-logs.txt        # KServe controller logs
└── warning-events.txt                # Warning/error events from all namespaces

Sharing with Red Hat Support

Package the collected data:

tar -czf rhaii-debug.tar.gz -C /tmp rhaii-on-xks-debug-*

Attach rhaii-debug.tar.gz to your support case.

Privacy and Security

The script collects: - Pod names, status, and logs - Resource definitions (CRDs, CRs) - Kubernetes events - Helm release metadata

The script does not collect: - Secret values (passwords, tokens, certificates) - Application data or model weights - Network traffic or request payloads

Review the output before sharing if your environment has specific data sensitivity requirements.

Prerequisites

  • kubectl configured with access to the cluster
  • Read access to the namespaces listed above

Troubleshooting the Script

If the script reports ERROR: Cannot connect to cluster:

# Verify kubeconfig
kubectl cluster-info

# Check current context
kubectl config current-context

If some sections show empty output, the corresponding component may not be deployed. This is expected if you haven't installed all components.