tundra-dome

What problem does it solve?

Standardizes bead/PR/issue flow across Gas Town → Tundra Dome, enabling consistent deployment, validation, and observability of the Tundra Dome stack (Kafka, Airflow, Datadog) in both local environments and KIND.

Core Features & Use Cases

• Lane routing, KPI snapshots, and Datadog dashboard validation
• Quick Tundra Dome rollout checks across hosts
• Kubernetes-based deployment bootstrap for local and KIND environments

Quick Start

Install prerequisites (bd, gt, kubectl, kind, docker, and the datadog-agent CLI if available).

To bootstrap the environment:

• Create namespaces: kubectl create ns tundra-dome && kubectl create ns datadog
• Create secrets: kubectl -n tundra-dome create secret generic tundra-dome-secrets --from-literal=DD_API_KEY=$DD_API_KEY --dry-run=client -o yaml | kubectl apply -f -
• kubectl -n datadog create secret generic tundra-dome-secrets --from-literal=DD_API_KEY=$DD_API_KEY --dry-run=client -o yaml | kubectl apply -f -
• kubectl apply -f infra/tundra-dome/tundra-dome.clean.yaml
• kubectl get pods -n tundra-dome -w Representative quick actions:
• Issue → bead sync (GitHub/Gitea): Set GITHUB_DRY_RUN=true to test; run daemon/kafka-dsm/github-issue-dispatcher.sh; then set GITHUB_DRY_RUN=false for live runs.
• Lane routing: Update labels to tundra-lane-* beads and adjust related config (airflow/dags and dispatcher scripts).
• KPI snapshot + DD checks: Run python3 daemon/kpi_snapshot.py and verify td_event_emitter metrics.
• Kafka health: Use td-kafka tools to verify topics and consumer groups.
• Launchd services: Restart services via launchctl and inspect logs.
• OpenCode/OpenRouter notes: Optional integration steps exist for model-based deployments.