CloudCostCompareDisaster recovery cost tiers: from backups to active-active
Short answer: disaster recovery (DR) is the plan for getting back after losing a data center or region. The faster you need to recover, the more you pay while nothing is wrong. Most systems fit backup and restore or pilot light. Reserve the expensive designs for the few that need them.
Key points
- Set recovery time (RTO) and data-loss limit (RPO) per system.
- Four designs: backup and restore, pilot light, warm standby, active-active.
- Cost rises from storage-only to about double.
- Replication traffic between regions is billed.
- Zones are not regions.
What are the two numbers to set first?
- Recovery time objective (RTO). How long you can be down.
- Recovery point objective (RPO). How much recent data you can lose.
Write both down for each system. A reporting tool and a payment system need different answers.
What are the four designs?
| Design | What runs in the second site | Cost shape | Recovery |
|---|---|---|---|
| Backup and restore | Nothing; copies of data are stored elsewhere | Storage and transfer only | Hours to days |
| Pilot light | A small core, such as a database replica; servers are off | Replica plus storage | Tens of minutes to hours |
| Warm standby | A scaled-down full copy that is running | A fraction of the production bill | Minutes |
| Active-active | A full copy serving live traffic | About double | Seconds, or none |
What does a worked example look like?
Made-up round numbers to show the method, not real prices. Production costs 2,000 a month: servers 800, database 900, other 300. Data is 500 GB and changes by 20 GB a day.
| Design | Extra monthly cost | Calculation |
|---|---|---|
| Backup and restore | 22 | 500 GB x 0.02 storage = 10.00, plus 600 GB replicated x 0.02 transfer = 12.00 |
| Pilot light | about 650 | database replica 900 x 0.6 (smaller size) = 540, plus storage 10, transfer 12, small extras 88 |
| Warm standby | about 1,000 | half-size servers 400, half-size database 450, storage and transfer 22, extras 128 |
| Active-active | about 2,000 | full duplicate, plus the cross-region traffic |
Moving from backup and restore to warm standby multiplies the extra cost by roughly forty-five. The question is whether an hour of downtime costs more than the difference. If an hour offline costs 500 and warm standby saves ten hours a year compared with pilot light, that is 5,000 a year for about 4,200 a year of extra standby cost. The numbers decide the tier, not preference.
What adds to the bill?
- Second copy of data. Storage in another region. See backup and retention costs.
- Replication traffic. Data sent between regions is billed. See data transfer costs.
- Standby compute and databases. Hourly charges while they sit ready.
- Load balancers and DNS in the second site.
- Tests. A plan never tested is a hope. Budget for drills.
Are zones the same as regions?
Spreading a system over several zones in one region protects against a data center failure and costs less than a second region. See multi-AZ costs. A regional outage needs the designs above.
Which mistakes are common?
- Paying for active-active on a system that can be down for an hour.
- Backups in the same region as the data.
- Never testing a restore.
- Forgetting that the second site needs the same secrets and configuration.
How do I choose?
Start from the cost of an hour of downtime for each system and compare it with the monthly cost of each tier. Most systems fit backup and restore or pilot light.
Decision checklist
- Do I have RTO and RPO for each system?
- Is a restore tested twice a year?
- Are backups outside the failure domain?
- Is the tier justified by the cost of downtime?
Next steps
Read multi-AZ and high availability costs for the cheaper first step, and managed database or self-hosted for how replication differs.
Frequently asked questions
What is disaster recovery in the cloud?
The plan for getting a system back after losing a data center or a whole region. It is set by two numbers: how long you can be down and how much recent data you can lose.
What are RTO and RPO?
Recovery time objective is how long you can be down. Recovery point objective is how much recent data you can lose. Set both for each system.
What are the main disaster recovery strategies?
Backup and restore, pilot light, warm standby and active-active. Cost and speed of recovery both rise in that order.
Is multi-zone the same as disaster recovery?
No. Several zones in one region protect against a data center failure. A regional outage needs a second region.
Do I need to test disaster recovery?
Yes. A plan never tested is a hope. Budget for drills.