Debugging Omnistrate Deployments

What problem does it solve?

This skill guides operators through a structured debugging workflow to identify root causes of Omnistrate deployment failures and noisy statuses while minimizing token usage.

Core Features & Use Cases

• Progressive two-phase workflow: Phase 1 gathers deployment status and related workflows; Phase 2 analyzes workflow events with targeted detail as needed.
• Application-level debugging assistance: When deployments are DEPLOYING with probe failures, the skill guides steps to collect logs, inspect resources, and isolate issues.
• Helm verification and broader context: Includes steps to verify Helm releases and optionally pull broader event data to triangulate root causes.
• Real-world scenarios: Use for failed deployments, unhealthy pods, startup probes failing, and unclear Helm states.

Quick Start

• Run the deployment status: omctl instance describe <instance-id> --deployment-status --output json
• List related workflows: omctl workflow list --service-id <id> --environment-id <id> --output json
• Inspect workflow events: omctl workflow events <workflow-id> --service-id <id> --environment-id <id> --output json
• If a resource needs detail: omctl workflow events <workflow-id> --service-id <id> --environment-id <id> --resource-key <name> --step-types Deployment --detail --output json
• For application-level debugging: omctl deployment-cell update-kubeconfig <cell-id> --kubeconfig /tmp/kubeconfig kubectl logs <pod-name> -n <instance-id> --kubeconfig /tmp/kubeconfig --tail=50