What's changed: In-scope coverage: added auto scaling (EC2 Auto Scaling), app integration (MQ/Step Functions/AppSync/AppFlow), migration/transfer (Application Migration Service/DMS/DataSync/Transfer Family/Snow Family), and hybrid connectivity (Direct Connect/Site-to-Site VPN/Client VPN/Transit Gateway/PrivateLink) across sections
2.3Backup and Disaster Recovery
Understand designing for disaster recovery: backups/snapshots, RTO and RPO, and multi-region failover (Route 53).
Resilience includes deciding how fast to recover (RTO) and how much data loss is acceptable (RPO), then designing backup/DR to match.
2.3.1Backup and DR elements
- Backups/snapshots: protect data with EBS snapshots, RDS automated backups, and S3 versioning / cross-region replication.
- RTO (recovery time objective): how fast to recover. RPO (recovery point objective): how far back you can recover = acceptable data loss.
- Four DR strategies: backup & restore (cheap/slow) → pilot light → warm standby → multi-site (costly/fast). Shorter RTO/RPO costs more.
- Multi-region: use Route 53 health checks + failover routing to switch to another region.
DR design starts by setting RTO (how fast to recover) and RPO (how much data loss is acceptable) as business requirements, then choosing a matching strategy. AWS’s four classic strategies trade cost against recovery speed: backup & restore (cheapest/slowest), pilot light (minimal always-on), warm standby (scaled-down always-on), multi-site/hot standby (full, fastest). Tighter RTO/RPO means more always-on resources and higher cost. For region-wide failures, replicate data via cross-region replication and use Route 53 health checks + failover routing to switch regions automatically.
| DR strategy | Always-on | RTO/RPO & cost |
|---|---|---|
| Backup & restore | None (restore on demand) | Long / cheapest |
| Pilot light | Minimal (e.g., DB only) | Medium |
| Warm standby | Scaled-down running | Short |
| Multi-site | Full running | Shortest / costliest |
Scenario: pick DR to fit requirements. If "a few hours of downtime and data loss back to the last backup" is acceptable, backup & restore suffices (low cost). If "recover within minutes with near-zero data loss" is required, use warm standby/multi-site + cross-region replication + Route 53 failover. Work backward from RTO/RPO to meet needs without over-investing.
Watch the mix-ups: (1) RTO = recovery speed (time) / RPO = how far back (data loss)—don’t swap. (2) Shorter RTO/RPO = higher cost (the four-strategy trade-off). (3) Route 53 = DNS-level failover (region switch). (4) Protect data with S3 cross-region replication/versioning.
Q. RTO vs RPO? RTO is "how fast to recover (time)"; RPO is "the point you can recover to = acceptable data loss." Q. How to choose a DR strategy? Balance tolerance for downtime/data loss (RTO/RPO) against cost, from backup & restore to multi-site. Q. Prepare for a region failure? Cross-region replication + Route 53 failover.
Common on SAA: RTO (recovery speed) vs RPO (acceptable data loss), the four DR strategies (backup & restore → multi-site) cost/speed trade-off, Route 53 failover routing, and S3 versioning/cross-region replication.
2.3.2Other key in-scope services (migration, transfer, hybrid connectivity)
For backup/DR and migration, services that move data and servers into AWS are central. AWS Application Migration Service lifts and shifts on-prem servers into AWS as-is; AWS Database Migration Service (DMS) migrates running databases with minimal downtime (including heterogeneous engines). AWS DataSync transfers large file sets fast and automatically between on-prem and AWS (S3/EFS/FSx); AWS Transfer Family moves files to S3/EFS over SFTP/FTPS/FTP for existing workflows. For limited bandwidth or offline bulk movement, use AWS Snow Family (physical devices to ship data).
For hybrid/multi-VPC connectivity, choose among these. AWS Direct Connect gives a dedicated line connecting on-prem to AWS with low latency and steady bandwidth (more consistent than VPN). AWS Site-to-Site VPN connects on-prem to a VPC over an encrypted tunnel across the internet (quick, low cost). AWS Client VPN gives secure remote access from individual user devices into a VPC. AWS Transit Gateway connects many VPCs and on-prem in a hub-and-spoke, and AWS PrivateLink provides private connectivity from a VPC to a service without traversing the internet (interface endpoints).
| What you want | Service |
|---|---|
| Migrate on-prem servers | AWS Application Migration Service |
| Migrate databases | AWS Database Migration Service |
| Fast bulk file transfer | AWS DataSync |
| SFTP/FTPS transfer | AWS Transfer Family |
| Offline bulk movement | AWS Snow Family |
| Dedicated-line connectivity | AWS Direct Connect |
| Encrypted tunnel to on-prem | AWS Site-to-Site VPN |
| Remote user access | AWS Client VPN |
| Hub for many VPCs | AWS Transit Gateway |
| Private connectivity to services | AWS PrivateLink |
2.3.3Section summary
- Set RTO (recovery speed) / RPO (acceptable data loss) as requirements and design matching DR
- Four strategies (backup & restore → multi-site) + Route 53 multi-region failover + S3 replication
Sign in to track progress — Log in.
Quick check
(just a quick review)Q1. Which metric expresses "how fast you must recover after a failure"?
Q2. Which metric expresses "how much data loss is acceptable"?
Q3. Which AWS service helps automatically fail over to another region?