SAP-C03 resilience and migration explained
Resilience, Migration, and Business Continuity is the fourth of the five SAP-C03 domains and carries 15% of scored content. It asks an architect to design workloads that survive failure at the level the business requires, to recover them when they do not, and to move existing workloads to AWS with a suitable strategy. SAP-C02 treated migration as a domain of its own; in SAP-C03 it is task 4.1 of this domain, next to disaster recovery and business continuity.
What the domain asks of an architect
- Recovery targets. RTO is how long a workload may be down; RPO is how much data may be lost. Both are set by the business and drive cost. See backup types and RTO/RPO.
- DR strategies. The AWS whitepaper on disaster recovery describes four, from cheapest and slowest to costliest and fastest: backup and restore, pilot light, warm standby, and multi-site active/active.
- Building blocks. Multi-AZ deployments, Auto Scaling, Elastic Load Balancing, Route 53 health checks and failover routing, Aurora Global Database, DynamoDB global tables, S3 replication, AWS Backup with cross-Region and cross-account copies and Vault Lock, and AWS Elastic Disaster Recovery.
- Migration. The common strategies are retire, retain, relocate, rehost, repurchase, replatform and refactor. Tools include AWS Application Migration Service for servers, AWS Database Migration Service and Schema Conversion Tool for databases, AWS DataSync and the Snow Family for data, and Migration Hub for tracking.
The three tasks in the AWS exam guide
| Task | What it covers |
|---|---|
| 4.1 Design migration strategies by using AWS frameworks | Discovery and portfolio assessment (dependencies, migration waves), rehosting with AWS Transform MGN (previously Application Migration Service), DMS and DataSync, replatforming to managed services, refactoring (microservices decomposition, strangler fig), modernizing mainframes, applications and databases |
| 4.2 Design disaster recovery architectures | DR to meet RTO, RPO and cost, continuous replication with Elastic Disaster Recovery, assessment and recovery procedures (Resilience Hub, Systems Manager runbooks), multi-Region failover (Route 53 health checks, Global Accelerator, Application Recovery Controller), DR testing, multi-cluster EKS and ECS |
| 4.3 Meet business continuity and resiliency requirements | Centralized and cross-account AWS Backup governance, failover tests and measured recovery times (Fault Injection Service), redundancy, fault tolerance and self-healing, ransomware and deletion protection (S3 Object Lock, EBS snapshot lock, Backup Vault Lock), chaos engineering with experiment templates and steady-state hypotheses |
New compared with SAP-C02, per the guide: multi-cluster container strategies for high availability and DR (task 4.2), and Resilience Hub and Fault Injection Service appear on the in-scope service list for the first time. The guide drops “migration-tool security” from the explicit list. Many SAP-C02 migration, DR and backup skills were consolidated into these three tasks.
Worked mini-scenario
An order system runs in one Region on Aurora and EC2. The business states an RTO of 15 minutes and an RPO of under a minute, with a budget well below doubling the environment.
- Backup and restore: cheap, but restores take hours and lose data since the last backup. It fails the RTO.
- Pilot light: core data replicated, compute off. Recovery needs time to start and scale, which may exceed 15 minutes.
- Warm standby: Aurora Global Database replicates across Regions and a scaled-down application tier runs in the second Region. Failover promotes the database and scales out. This fits both targets at a fraction of full duplication.
- Multi-site active/active: meets the targets with the best margin but costs the most and needs data-conflict design.
Warm standby is the best fit, and the architect should validate it. A Fault Injection experiment or a failover drill shows whether the 15 minutes is real, and Application Recovery Controller can make the failover decision controlled rather than ad hoc. The trade-off is ongoing cost for the standby tier.
Easy-to-confuse pairs
| Pair | The distinction |
|---|---|
| RTO vs RPO | Time to recover vs data that may be lost |
| High availability vs disaster recovery | Surviving a component or AZ failure vs surviving loss of a Region |
| Multi-AZ vs read replica | Synchronous standby for failover vs asynchronous copy for reads |
| Aurora Global Database vs cross-Region read replica | Dedicated storage-level replication, fast managed failover vs binlog-based replica |
| Rehost vs replatform vs refactor | Lift as is, small optimisations, redesign |
| Application Migration Service vs DMS | Servers vs databases |
| Snowball vs DataSync | Offline physical transfer vs online managed transfer |
Exam-style pitfall
Treating a backup as recovery. A backup that has never been restored, or that lives in the same account and Region as the failure, does not meet an RTO. Equally, choosing multi-site for an RTO of hours overspends: the cheapest strategy that meets both targets is the answer.
Practice questions
Question 1. A company has a 20 TB on-premises Oracle database and wants to move to Amazon Aurora PostgreSQL with minimal downtime and with schema and code conversion help. Which combination fits best?
- A. AWS DataSync to copy the data files
- B. AWS Snowball Edge shipping database files
- C. AWS Application Migration Service to rehost the database server
- D. AWS Schema Conversion Tool with AWS Database Migration Service replication
Show answer
Answer: D
AWS Schema Conversion Tool converts the schema and code for a heterogeneous move, and AWS Database Migration Service performs the full load and ongoing replication so cutover downtime is short. DataSync moves files, Snowball alone does not convert or replicate changes, and Application Migration Service rehosts servers rather than changing database engines.
Want more questions like this? Full SAP-C03 practice tests →
Question 2. A workload has an RTO of 4 hours and an RPO of 24 hours. It is not customer facing, and cost must be kept as low as possible. Which DR strategy should be chosen?
- A. Multi-site active/active in two Regions
- B. Backup and restore with cross-Region backup copies
- C. Warm standby with a scaled-down copy running
- D. Pilot light with continuous replication
Show answer
Answer: B
Backup and restore, using AWS Backup with copies in a second Region, meets a 24-hour RPO and 4-hour RTO at the lowest cost because no standby infrastructure runs. Warm standby and multi-site run infrastructure continuously and exceed the requirements, and pilot light keeps replicated data and idle environment components that the targets do not require.
Want more questions like this? Full SAP-C03 practice tests →
Question 3. A team must prove that an application meets its stated recovery objectives and wants to find unexpected dependencies by injecting failures such as terminated instances and added network latency in a controlled way. Which service helps?
- A. AWS CloudTrail
- B. AWS Backup
- C. AWS Fault Injection Service
- D. AWS Trusted Advisor
Show answer
Answer: C
AWS Fault Injection Service runs controlled fault experiments with stop conditions, which exposes hidden dependencies and tests recovery behaviour. CloudTrail records API activity, Trusted Advisor gives general check recommendations, and AWS Backup stores recovery points but does not inject failures.
Want more questions like this? Full SAP-C03 practice tests →