Disaster recovery tiers
Pilot light, warm standby and hot standby side by side — the same application at three price points of RTO and RPO.
What this design does
Disaster recovery is a slider, not a switch, and this template shows the three classic positions next to each other. Pilot light keeps only data replicated — cheap, but recovery means booting everything (hours). Warm standby runs a minimal copy that scales up on failover (minutes). Hot standby runs full capacity with near-synchronous data (seconds, and the biggest bill). Pick by answering two questions honestly: how much downtime can the business survive (RTO), and how much data can it afford to lose (RPO)?
The patterns it composes
- DR tiers (pilot light / warm / hot) — three standby postures trading cost against recovery speed
- RTO / RPO-driven topology — the tier is chosen by recovery-time and data-loss budgets, not by fashion
- Async vs sync replication — pilot/warm ship changes asynchronously; hot pays the latency for near-sync
- Region evacuation runbook — failover is a rehearsed procedure with an owner — not a wiki page
- Backup-and-restore (tier 0) — beneath every tier: real backups, actually restored in drills
Import it, attach your own numbers, and verify the design against failures — or read the engineering notes.