Backup and restore tier labels — farm stack¶

Purpose¶

Durable labels for Tier 0–3 recovery expectations used across farm records, volumes, and platform rebuilds—so pages can say “Tier 2 path” without pasting the tier table.

Scope¶

• In scope: What each tier means operationally; pointers to RPO/RTO as operator-filled worksheets.
• Out of scope: Numeric SLAs invented by the wiki; your drill dates (those belong in runbooks / evidence).

Knowns¶

• Tier 0 is live state (no restore).
• Tier 1 is typically file/volume rollback (restic/Longhorn-class).
• Tier 2 is logical DB / app restore (e.g. pg_dump class).
• Tier 3 is platform DR (cluster/storage rebuild ordering).

Assumptions¶

• Tier 3 without a Tier 2-class path for the same app data can yield an “empty healthy” system—see canonical warning in restore tiers analysis.

Related analyses¶

• Disaster recovery decision rules — farm edge stack — when to apply each tier in an incident
• Restore and recovery tiers — homelab farm systems
• Backup and disaster recovery package — smart farm stack
• Runbook — backup validation and recovery drill
• Kubernetes platform backup / DR — Pi fleet, k3s, Longhorn

Related source notes¶

• Backup / DR — official documentation cluster

Outstanding unknowns¶

• Dated restore drill outcomes per tier (evidence gap until runbooks are exercised).