Amazon AWS Certified Solutions Architect - Associate SAA-C03 Practice Test 38
Question 370
An RDS Multi-AZ DB instance deployment has one passive standby; read-only queries overwhelm the writer. What additional architecture is needed?
- Add appropriate RDS read replicas or choose a Multi-AZ DB cluster offering readable standbys when supported.
- A traditional Multi-AZ DB instance standby is for failover, not readable query offload.
- Deploy a supported RDS Multi-AZ configuration for the database engine.
- Use an RDS Multi-AZ DB cluster with two readable standby DB instances.
Explanation
Correct Answer
A:A traditional Multi-AZ DB instance standby does not serve read queries, so HA alone does not scale reads. Direct test reads to the reader endpoint and observe writer load.
Incorrect Answers
B:Read scaling requires a supported read replica or a different readable Multi-AZ cluster architecture. Add appropriate RDS read replicas or choose a Multi-AZ DB cluster offering readable standbys when supported. Direct test reads to the reader endpoint and observe writer load.
C:RDS provisions standby capacity in a separate AZ and manages failover on supported configurations. Add appropriate RDS read replicas or choose a Multi-AZ DB cluster offering readable standbys when supported. Direct test reads to the reader endpoint and observe writer load.
D:Unlike traditional one-standby Multi-AZ instances, Multi-AZ DB clusters can expose reader capacity. Add appropriate RDS read replicas or choose a Multi-AZ DB cluster offering readable standbys when supported. Direct test reads to the reader endpoint and observe writer load.
Question 371
A team needs both in-Region failover and readable standby DB instances across three AZs in an engine that supports the feature. Which RDS configuration fits?
- Use an RDS Multi-AZ DB cluster with two readable standby DB instances.
- Add an ordinary read replica but leave the traditional Multi-AZ standby unreadable while claiming two readable standbys across three AZs.
- A traditional Multi-AZ DB instance standby is for failover, not readable query offload.
- Deploy a supported RDS Multi-AZ configuration for the database engine.
Explanation
Correct Answer
A:Unlike traditional one-standby Multi-AZ instances, Multi-AZ DB clusters can expose reader capacity. Query writer and reader endpoints and simulate a primary outage.
Incorrect Answers
B:An additional read replica is not automatically equivalent to a Multi-AZ DB cluster with two readable standby members. The explicit requirement calls for a supported Multi-AZ DB cluster architecture.
C:Read scaling requires a supported read replica or a different readable Multi-AZ cluster architecture. Use an RDS Multi-AZ DB cluster with two readable standby DB instances. Query writer and reader endpoints and simulate a primary outage.
D:RDS provisions standby capacity in a separate AZ and manages failover on supported configurations. Use an RDS Multi-AZ DB cluster with two readable standby DB instances. Query writer and reader endpoints and simulate a primary outage.
Question 372
An application stores an RDS writer instance's old private IP and fails after automatic failover. What should it reference?
- Configure RDS automated backups with a justified retention period.
- Use an RDS Multi-AZ DB cluster with two readable standby DB instances.
- Connect through the supported RDS DNS endpoint and implement connection retries.
- Deploy a supported RDS Multi-AZ configuration for the database engine.
Explanation
Correct Answer
C:RDS endpoints can move to replacement instances; hardcoded addresses defeat failover. Perform a failover and verify DNS resolution and application reconnection.
Incorrect Answers
A:Automated backups and transaction logs support PITR within the retention bounds of supported engines. Connect through the supported RDS DNS endpoint and implement connection retries. Perform a failover and verify DNS resolution and application reconnection.
B:Unlike traditional one-standby Multi-AZ instances, Multi-AZ DB clusters can expose reader capacity. Connect through the supported RDS DNS endpoint and implement connection retries. Perform a failover and verify DNS resolution and application reconnection.
D:RDS provisions standby capacity in a separate AZ and manages failover on supported configurations. Connect through the supported RDS DNS endpoint and implement connection retries. Perform a failover and verify DNS resolution and application reconnection.
Question 373
A transaction database needs an automated point-in-time restore within the configured retention window. Which feature should be enabled?
- Connect through the supported RDS DNS endpoint and implement connection retries.
- Configure RDS automated backups with a justified retention period.
- Deploy a supported RDS Multi-AZ configuration for the database engine.
- Copy suitable RDS snapshots to the recovery Region and validate required encryption permissions.
Explanation
Correct Answer
B:Automated backups and transaction logs support PITR within the retention bounds of supported engines. Restore a staging copy to a timestamp preceding a test update.
Incorrect Answers
A:RDS endpoints can move to replacement instances; hardcoded addresses defeat failover. Configure RDS automated backups with a justified retention period. Restore a staging copy to a timestamp preceding a test update.
C:RDS provisions standby capacity in a separate AZ and manages failover on supported configurations. Configure RDS automated backups with a justified retention period. Restore a staging copy to a timestamp preceding a test update.
D:Multi-AZ alone does not place recoverable backups in another Region. Configure RDS automated backups with a justified retention period. Restore a staging copy to a timestamp preceding a test update.
Question 374
An RDS standby in another AZ is mistakenly expected to handle reports, but the deployment is a traditional Multi-AZ DB instance. What distinction matters?
- Use an RDS Multi-AZ DB cluster with two readable standby DB instances.
- Add appropriate RDS read replicas or choose a Multi-AZ DB cluster offering readable standbys when supported.
- Connect through the supported RDS DNS endpoint and implement connection retries.
- A traditional Multi-AZ DB instance standby is for failover, not readable query offload.
Explanation
Correct Answer
D:Read scaling requires a supported read replica or a different readable Multi-AZ cluster architecture. Attempt a read through supported endpoints and inspect database roles.
Incorrect Answers
A:Unlike traditional one-standby Multi-AZ instances, Multi-AZ DB clusters can expose reader capacity. A traditional Multi-AZ DB instance standby is for failover, not readable query offload. Attempt a read through supported endpoints and inspect database roles.
B:A traditional Multi-AZ DB instance standby does not serve read queries, so HA alone does not scale reads. A traditional Multi-AZ DB instance standby is for failover, not readable query offload. Attempt a read through supported endpoints and inspect database roles.
C:RDS endpoints can move to replacement instances; hardcoded addresses defeat failover. Read scaling requires a supported read replica or a different readable Multi-AZ cluster architecture.
Question 375
RDS reports failovers but the application takes too long to recover due to stale connections. What should the app implement?
- Use bounded reconnect/retry logic and avoid pinning to individual DB hosts.
- A traditional Multi-AZ DB instance standby is for failover, not readable query offload.
- Connect through the supported RDS DNS endpoint and implement connection retries.
- Deploy a supported RDS Multi-AZ configuration for the database engine.
Explanation
Correct Answer
A:Managed failover still requires clients to resolve current endpoints and reestablish sessions. Force failover and measure application error recovery after reconnect changes.
Incorrect Answers
B:Read scaling requires a supported read replica or a different readable Multi-AZ cluster architecture. Use bounded reconnect/retry logic and avoid pinning to individual DB hosts. Force failover and measure application error recovery after reconnect changes.
C:RDS endpoints can move to replacement instances; hardcoded addresses defeat failover. Use bounded reconnect/retry logic and avoid pinning to individual DB hosts. Force failover and measure application error recovery after reconnect changes.
D:RDS provisions standby capacity in a separate AZ and manages failover on supported configurations. Use bounded reconnect/retry logic and avoid pinning to individual DB hosts. Force failover and measure application error recovery after reconnect changes.
Question 376
A finance database's backup snapshot is needed in a second AWS Region for disaster recovery. What is the relevant additional action?
- Deploy a supported RDS Multi-AZ configuration for the database engine.
- Configure RDS automated backups with a justified retention period.
- Add appropriate RDS read replicas or choose a Multi-AZ DB cluster offering readable standbys when supported.
- Copy suitable RDS snapshots to the recovery Region and validate required encryption permissions.
Explanation
Correct Answer
D:Multi-AZ alone does not place recoverable backups in another Region. Restore a test snapshot in the destination Region under the recovery account.
Incorrect Answers
A:RDS provisions standby capacity in a separate AZ and manages failover on supported configurations. Copy suitable RDS snapshots to the recovery Region and validate required encryption permissions. Restore a test snapshot in the destination Region under the recovery account.
B:Automated backups and transaction logs support PITR within the retention bounds of supported engines. Copy suitable RDS snapshots to the recovery Region and validate required encryption permissions. Restore a test snapshot in the destination Region under the recovery account.
C:A traditional Multi-AZ DB instance standby does not serve read queries, so HA alone does not scale reads. Copy suitable RDS snapshots to the recovery Region and validate required encryption permissions. Restore a test snapshot in the destination Region under the recovery account.
Question 377
An Aurora cluster writer fails but healthy replicas already exist in different AZs. Which mechanism promotes service continuity?
- Use Aurora Global Database with primary-Region writes and secondary-Region reads.
- Route read-only queries to the Aurora reader endpoint with adequate reader instances.
- Pin database writes to the old writer instance endpoint and bypass Aurora managed writer-role routing after failure.
- Use Aurora's managed failover to an eligible reader and connect through the writer cluster endpoint.
Explanation
Correct Answer
D:Aurora promotes an appropriate replica and the cluster endpoint tracks the current writer. Induce a controlled failover and verify reconnection through the writer endpoint.
Incorrect Answers
A:Aurora's global topology separates the writer Region from secondary reader Regions under normal operation. Use Aurora's managed failover to an eligible reader and connect through the writer cluster endpoint. Induce a controlled failover and verify reconnection through the writer endpoint.
B:The reader endpoint spreads eligible reads among Aurora replicas without changing writer access. Use Aurora's managed failover to an eligible reader and connect through the writer cluster endpoint. Induce a controlled failover and verify reconnection through the writer endpoint.
C:An old instance-specific writer hostname may no longer represent the promoted writer after failover. The Aurora managed failover path plus writer cluster endpoint preserves a route to the new writer.
Question 378
A report workload must stop overloading Aurora's writer while retaining transactional writes. What endpoint should reports use?
- Configure read workloads to use the reader endpoint and verify reader-capable instances exist.
- Route read-only queries to the Aurora reader endpoint with adequate reader instances.
- Use Aurora's managed failover to an eligible reader and connect through the writer cluster endpoint.
- Use Aurora Global Database with primary-Region writes and secondary-Region reads.
Explanation
Correct Answer
B:The reader endpoint spreads eligible reads among Aurora replicas without changing writer access. Run reporting queries through reader endpoint and measure writer CPU.
Incorrect Answers
A:A cluster endpoint for writing does not automatically distribute reads to replicas. Route read-only queries to the Aurora reader endpoint with adequate reader instances. Run reporting queries through reader endpoint and measure writer CPU.
C:Aurora promotes an appropriate replica and the cluster endpoint tracks the current writer. Route read-only queries to the Aurora reader endpoint with adequate reader instances. Run reporting queries through reader endpoint and measure writer CPU.
D:Aurora's global topology separates the writer Region from secondary reader Regions under normal operation. Route read-only queries to the Aurora reader endpoint with adequate reader instances. Run reporting queries through reader endpoint and measure writer CPU.
Question 379
A database must replicate across Regions with fast recovery and global readers. Which feature is designed for that role?
- Secondary-Region reads can reflect replication lag and may not be immediately current.
- Route read-only queries to the Aurora reader endpoint with adequate reader instances.
- Use an Aurora cluster with replicas in multiple Availability Zones but no cross-Region replication.
- Use Aurora Global Database with a primary and one or more secondary Regions.
Explanation
Correct Answer
D:Global Database replicates storage data across Regions for low-latency reads and cross-Region recovery. Query a secondary Region and perform a documented DR switchover test.
Incorrect Answers
A:Cross-Region replication is asynchronous from primary to secondary, so read-after-write consistency is not automatic globally. Use Aurora Global Database with a primary and one or more secondary Regions. Query a secondary Region and perform a documented DR switchover test.
B:The reader endpoint spreads eligible reads among Aurora replicas without changing writer access. Use Aurora Global Database with a primary and one or more secondary Regions. Query a secondary Region and perform a documented DR switchover test.
C:Replicas in several AZs protect the database within one Region, but cannot provide the stated global readers and recovery in another Region. Aurora Global Database explicitly supports a primary Region and secondary Regions.
Question 380
A globally distributed application needs strong writes in one Region but low-latency local reads in others. Which database layout fits?
- Use Aurora's managed failover to an eligible reader and connect through the writer cluster endpoint.
- Use Aurora Global Database with primary-Region writes and secondary-Region reads.
- Allow applications in every secondary Aurora Global Database Region to issue ordinary writes directly to the read-only regional clusters.
- Route read-only queries to the Aurora reader endpoint with adequate reader instances.
Explanation
Correct Answer
B:Aurora's global topology separates the writer Region from secondary reader Regions under normal operation. Verify writes enter the primary and secondary readers reflect replicated updates.
Incorrect Answers
A:Aurora promotes an appropriate replica and the cluster endpoint tracks the current writer. Use Aurora Global Database with primary-Region writes and secondary-Region reads. Verify writes enter the primary and secondary readers reflect replicated updates.
C:Standard Aurora Global Database secondary clusters are read-only until an authorized switchover/failover changes roles; treating every Region as a concurrent writer is not the requested single-writer design.
D:The reader endpoint spreads eligible reads among Aurora replicas without changing writer access. Use Aurora Global Database with primary-Region writes and secondary-Region reads. Verify writes enter the primary and secondary readers reflect replicated updates.