# Product Guidelines - JuiceNavidrome Infrastructure

## Vision
To provide a rock-solid, self-healing, and highly available deployment model for Navidrome, treating the infrastructure as a first-class product.

## Operational Principles
- **Infrastructure as Code:** All cluster configurations, job definitions, and container builds must be versioned and managed through Git.
- **Failover-First Design:** The primary measure of success is the system's ability to automatically recover from node failures without manual intervention.
- **Minimal State on Host:** Local host data should be limited to necessary LiteFS state and caches; all critical application data must be replicated or reside on shared storage.

## Automation Standards
- **Automated Rebuilds:** Any change to the base configuration or Dockerfile must trigger an automatic build and push to the local registry via Gitea Actions.
- **Safe Rollouts:** Use sequential update strategies (max_parallel=1) to ensure the cluster remains healthy during version transitions.

## Reliability & Monitoring
- **Health-Aware Orchestration:** Leverage Consul health checks to provide Traefik with accurate routing information, ensuring traffic only hits ready nodes.
- **Consistent Initialization:** Use clean bootstrap procedures to avoid common distributed system pitfalls like stale cluster IDs or locking conflicts.