Resilience, Migration, and Business Continuity15% · 4 topics

Q1. Which statement about RPO and RTO is incorrect?

1. RPO is the target for how long it takes from a failure until the system recovers
2. Generally, the shorter the RPO and RTO, the higher the cost of the DR mechanism
3. RPO expresses acceptable data loss as the point in time you must be able to go back to
4. RTO is the target for the maximum acceptable time from a service outage until recovery

Q2. During migration with DMS, CDC latency keeps increasing and cannot catch up. What is the most appropriate next step?

1. Switch LOB columns to full LOB mode to prevent missing data
2. Enlarge the target DB instance to speed up writes
3. Restart the task from full load to reset the CDC start position
4. Check the replication instance's CPU, memory and more, and move to a larger class if needed

Q1. In a DR drill switching to the Osaka Region, EC2 launches failed partway. The cause was the Region's service quotas. What is the most appropriate next step?

1. Operate so that when a disaster happens, you contact AWS Support then to raise quotas
2. Change DR instance types to the smallest so they can launch within quotas
3. Check the quotas needed for production scale and request increases in the DR Region in advance
4. Use another cloud with no quota limits as the DR site and stop doing drills

Q2. If DynamoDB table data is corrupted by a staff member's mistaken bulk update, you want to restore it to the state just before. Which method is most appropriate?

1. Replicate to another Region with global tables and restore from the replica of the corrupted data
2. Enable point-in-time recovery (PITR) to restore to any point within the retention period
3. Disable DynamoDB Streams so updates do not propagate to other tables
4. Increase table read/write capacity so updates finish as fast as possible

Q1. You want to ensure newly created EBS and RDS resources are not missed by backups. Which method has the least operational overhead?

1. Backups can be taken when a failure occurs, so do not include new resources
2. Assign an AWS Backup backup plan by tag so resources with a defined tag are included automatically
3. Each time a resource is created, have staff add it to a backup ledger and take snapshots manually
4. Once a month, export a list of resources in all accounts and visually check for missed backups

Q2. A DR design exists but RTO compliance is unproven. What comes next?

1. Remove alarm notification recipients
2. Measure failover and data consistency in a recovery drill
3. Only write RTO numbers on a diagram
4. Keep production backups indefinitely

Q1. In a production-like environment, you want to verify that automatic recovery mechanisms and procedures really work when an AZ fails or instances stop. Which method is most appropriate?

1. Plan experiments that deliberately inject faults with AWS FIS, define stop conditions and observe the impact
2. Assume failures will not happen and consider reviewing design documents sufficient for automatic recovery
3. In production, have staff manually stop instances one after another without notice to check
4. Judge that the mechanisms work correctly because no failure has ever occurred

Q2. You want to eliminate single points of failure (SPOFs) in an existing architecture. Private subnets in three AZs all reach the internet through a single NAT gateway in AZ-a. Which measure is the most suitable?

1. Make Direct Connect redundant at another location
2. Move the files to Amazon EFS
3. Place a NAT gateway in each AZ
4. Make RDS a Multi-AZ deployment