Cloud Computing Tutorial · Integration and Operations · Lesson 22 of 24
Reliability and Disaster Recovery
RTO and RPO, the four disaster-recovery strategies from backup-and-restore to multi-site, and testing restores.
RTO and RPO
- RTO (recovery time objective): how long the service may be down.
- RPO (recovery point objective): how much data (in time) may be lost.
The business sets them per system; they decide the strategy and the cost.
Strategies
| Strategy | How | RTO / RPO | Cost |
|---|---|---|---|
| Backup and restore | Backups copied to another region; rebuild from code when needed | Hours / hours | Low |
| Pilot light | Data replicated continuously; core infrastructure minimal in the second region | Tens of minutes / minutes | Medium |
| Warm standby | A scaled-down full copy running in the second region | Minutes / seconds to minutes | Higher |
| Multi-site active-active | Full capacity in several regions serving traffic | Near zero / near zero | Highest |
strategies = { # (RTO minutes, RPO minutes, relative monthly cost)
"backup and restore": (480, 240, 1.05),
"pilot light": (45, 5, 1.2),
"warm standby": (10, 1, 1.5),
"multi-site": (1, 0.1, 2.1),
}
required_rto, required_rpo = 60, 15
for name, (rto, rpo, cost) in strategies.items():
ok = rto <= required_rto and rpo <= required_rpo
print(f"{name:20} RTO {rto:5} min RPO {rpo:5} min cost x{cost:<4} {'meets target' if ok else ''}")
cheapest = min((c, n) for n, (r, p, c) in strategies.items() if r <= required_rto and p <= required_rpo)
print("cheapest strategy meeting RTO 60 / RPO 15:", cheapest[1])
Output
backup and restore RTO 480 min RPO 240 min cost x1.05 pilot light RTO 45 min RPO 5 min cost x1.2 meets target warm standby RTO 10 min RPO 1 min cost x1.5 meets target multi-site RTO 1 min RPO 0.1 min cost x2.1 meets target cheapest strategy meeting RTO 60 / RPO 15: pilot light
Rules that make DR real
- Infrastructure as code, so a region can be rebuilt quickly.
- Backups in another account and region, immutable (AWS Backup Vault Lock, Azure immutable vaults, Google backup vaults) so attackers or mistakes can't delete them.
- Test restores and failovers regularly; an untested backup is a hope, not a plan.
Test yourself
A system may lose at most 5 minutes of data. Which objective is that?
RTORPOSLAMTTR
Show answer
Answer: RPO
The recovery point objective is about data loss; RTO is about downtime.