Amazon AWS Certified Solutions Architect - Associate SAA-C03 Bundle

Store large objects in S3 and retain references in DynamoDB would help if every item includes huge rarely used blobs and bloats read capacity. Splitting large blobs can reduce item size and read cost. That is not the symptom described here: provisioned throughput remains high overnight despite low traffic. The matching response is Set DynamoDB target-tracking auto scaling or use on-demand based on patterns, because adaptive capacity mode reduces idle provisioned spend.

For the alternative failure the app needs strongly consistent reads after every write, the control Use DynamoDB consistent reads where required instead of relying on stale caches is relevant: strong consistency may be needed for immediate correctness. Here the evidence instead is provisioned throughput remains high overnight despite low traffic. The selected technical mechanism is Set DynamoDB target-tracking auto scaling or use on-demand based on patterns: adaptive capacity mode reduces idle provisioned spend.

Use DynamoDB TTL and bounded retention workflows solves a different operational problem where unused old events remain in DynamoDB tables forever. TTL removes eligible expired items. It does not directly address provisioned throughput remains high overnight despite low traffic. The case calls for Set DynamoDB target-tracking auto scaling or use on-demand based on patterns, since adaptive capacity mode reduces idle provisioned spend.

Use Multi-AZ within one Region rather than duplicating full multi-region writers would help if a reporting database must remain available during one AZ failure but not region loss. AZ redundancy can satisfy narrower availability objectives. That is not the symptom described here: a low-value single-region test database is replicated to three Regions without requirements. The matching response is Remove unnecessary geographic replicas after reviewing DR policy, because regional replicas incur ongoing storage/compute costs.

For the alternative failure a global write-heavy retail system demands regional reads and writes, the control Use DynamoDB global tables if conflict/replication tradeoffs fit is relevant: business availability can justify multi-region replication costs. Here the evidence instead is a low-value single-region test database is replicated to three Regions without requirements. The selected technical mechanism is Remove unnecessary geographic replicas after reviewing DR policy: regional replicas incur ongoing storage/compute costs.

The competing option Establish recovery objectives before selecting regional data replication would be useful when an application has no defined RPO/RTO but pays for the most expensive database DR pattern. Cost must be justified by required recovery outcomes. The present issue is a low-value single-region test database is replicated to three Regions without requirements; it is more directly addressed by Remove unnecessary geographic replicas after reviewing DR policy. Regional replicas incur ongoing storage/compute costs.

Use Multi-AZ within one Region rather than duplicating full multi-region writers solves a different operational problem where a reporting database must remain available during one AZ failure but not region loss. AZ redundancy can satisfy narrower availability objectives. It does not directly address a global write-heavy retail system demands regional reads and writes. The case calls for Use DynamoDB global tables if conflict/replication tradeoffs fit, since business availability can justify multi-region replication costs.

The competing option Establish recovery objectives before selecting regional data replication would be useful when an application has no defined RPO/RTO but pays for the most expensive database DR pattern. Cost must be justified by required recovery outcomes. The present issue is a global write-heavy retail system demands regional reads and writes; it is more directly addressed by Use DynamoDB global tables if conflict/replication tradeoffs fit. Business availability can justify multi-region replication costs.

The option Scale reporting readers or use scheduled extracts under acceptable freshness addresses a separate condition: an infrequent report uses a continuously running large Aurora reader. Scheduled analytics can reduce idle reader cost. The actual requirement is a global write-heavy retail system demands regional reads and writes. Recommended corrective action: Use DynamoDB global tables if conflict/replication tradeoffs fit. Business availability can justify multi-region replication costs.

Scale reporting readers or use scheduled extracts under acceptable freshness solves a different operational problem where an infrequent report uses a continuously running large Aurora reader. Scheduled analytics can reduce idle reader cost. It does not directly address a reporting database must remain available during one AZ failure but not region loss. The case calls for Use Multi-AZ within one Region rather than duplicating full multi-region writers, since aZ redundancy can satisfy narrower availability objectives.

The competing option Use DynamoDB global tables if conflict/replication tradeoffs fit would be useful when a global write-heavy retail system demands regional reads and writes. Business availability can justify multi-region replication costs. The present issue is a reporting database must remain available during one AZ failure but not region loss; it is more directly addressed by Use Multi-AZ within one Region rather than duplicating full multi-region writers. AZ redundancy can satisfy narrower availability objectives.

Establish recovery objectives before selecting regional data replication would help if an application has no defined RPO/RTO but pays for the most expensive database DR pattern. Cost must be justified by required recovery outcomes. That is not the symptom described here: a reporting database must remain available during one AZ failure but not region loss. The matching response is Use Multi-AZ within one Region rather than duplicating full multi-region writers, because aZ redundancy can satisfy narrower availability objectives.

The competing option Remove unnecessary geographic replicas after reviewing DR policy would be useful when a low-value single-region test database is replicated to three Regions without requirements. Regional replicas incur ongoing storage/compute costs. The present issue is an infrequent report uses a continuously running large Aurora reader; it is more directly addressed by Scale reporting readers or use scheduled extracts under acceptable freshness. Scheduled analytics can reduce idle reader cost.

The option Establish recovery objectives before selecting regional data replication addresses a separate condition: an application has no defined RPO/RTO but pays for the most expensive database DR pattern. Cost must be justified by required recovery outcomes. The actual requirement is an infrequent report uses a continuously running large Aurora reader. Recommended corrective action: Scale reporting readers or use scheduled extracts under acceptable freshness. Scheduled analytics can reduce idle reader cost.

For the alternative failure a global write-heavy retail system demands regional reads and writes, the control Use DynamoDB global tables if conflict/replication tradeoffs fit is relevant: business availability can justify multi-region replication costs. Here the evidence instead is an infrequent report uses a continuously running large Aurora reader. The selected technical mechanism is Scale reporting readers or use scheduled extracts under acceptable freshness: scheduled analytics can reduce idle reader cost.

The option Remove unnecessary geographic replicas after reviewing DR policy addresses a separate condition: a low-value single-region test database is replicated to three Regions without requirements. Regional replicas incur ongoing storage/compute costs. The actual requirement is an application has no defined RPO/RTO but pays for the most expensive database DR pattern. Recommended corrective action: Establish recovery objectives before selecting regional data replication. Cost must be justified by required recovery outcomes.

For the alternative failure a global write-heavy retail system demands regional reads and writes, the control Use DynamoDB global tables if conflict/replication tradeoffs fit is relevant: business availability can justify multi-region replication costs. Here the evidence instead is an application has no defined RPO/RTO but pays for the most expensive database DR pattern. The selected technical mechanism is Establish recovery objectives before selecting regional data replication: cost must be justified by required recovery outcomes.

Use Multi-AZ within one Region rather than duplicating full multi-region writers solves a different operational problem where a reporting database must remain available during one AZ failure but not region loss. AZ redundancy can satisfy narrower availability objectives. It does not directly address an application has no defined RPO/RTO but pays for the most expensive database DR pattern. The case calls for Establish recovery objectives before selecting regional data replication, since cost must be justified by required recovery outcomes.

Keep dedicated databases/accounts according to compliance rather than pooling would help if tenants require physical data separation by legal contract. Isolation requirements override cost optimization. That is not the symptom described here: each of 100 small customers has a dedicated mostly idle RDS instance. The matching response is Consider shared multitenant schema/database with rigorously enforced tenant isolation, because pooling may reduce small-instance overhead when isolation requirements allow.

Enforce tenant-scoped database authorization and key-based access solves a different operational problem where a shared table filters tenants only in app memory after querying all records. Safe multitenancy requires robust data-level boundaries. It does not directly address each of 100 small customers has a dedicated mostly idle RDS instance. The case calls for Consider shared multitenant schema/database with rigorously enforced tenant isolation, since pooling may reduce small-instance overhead when isolation requirements allow.

The competing option Consider Aurora Serverless v2 or DynamoDB on-demand where workload semantics fit would be useful when customers' usage swings unpredictably throughout the day. Elastic pricing aligns capacity with variable demand. The present issue is each of 100 small customers has a dedicated mostly idle RDS instance; it is more directly addressed by Consider shared multitenant schema/database with rigorously enforced tenant isolation. Pooling may reduce small-instance overhead when isolation requirements allow.

For the alternative failure customers' usage swings unpredictably throughout the day, the control Consider Aurora Serverless v2 or DynamoDB on-demand where workload semantics fit is relevant: elastic pricing aligns capacity with variable demand. Here the evidence instead is tenants require physical data separation by legal contract. The selected technical mechanism is Keep dedicated databases/accounts according to compliance rather than pooling: isolation requirements override cost optimization.

The competing option Implement compliant tiering and expiration for historical records would be useful when storage costs rise because tenant audit histories are retained indefinitely. Data lifecycle policies control unbounded growth. The present issue is tenants require physical data separation by legal contract; it is more directly addressed by Keep dedicated databases/accounts according to compliance rather than pooling. Isolation requirements override cost optimization.

The option Consider shared multitenant schema/database with rigorously enforced tenant isolation addresses a separate condition: each of 100 small customers has a dedicated mostly idle RDS instance. Pooling may reduce small-instance overhead when isolation requirements allow. The actual requirement is tenants require physical data separation by legal contract. Recommended corrective action: Keep dedicated databases/accounts according to compliance rather than pooling. Isolation requirements override cost optimization.

Consider shared multitenant schema/database with rigorously enforced tenant isolation solves a different operational problem where each of 100 small customers has a dedicated mostly idle RDS instance. Pooling may reduce small-instance overhead when isolation requirements allow. It does not directly address a shared table filters tenants only in app memory after querying all records. The case calls for Enforce tenant-scoped database authorization and key-based access, since safe multitenancy requires robust data-level boundaries.

The competing option Consider Aurora Serverless v2 or DynamoDB on-demand where workload semantics fit would be useful when customers' usage swings unpredictably throughout the day. Elastic pricing aligns capacity with variable demand. The present issue is a shared table filters tenants only in app memory after querying all records; it is more directly addressed by Enforce tenant-scoped database authorization and key-based access. Safe multitenancy requires robust data-level boundaries.

Implement compliant tiering and expiration for historical records would help if storage costs rise because tenant audit histories are retained indefinitely. Data lifecycle policies control unbounded growth. That is not the symptom described here: a shared table filters tenants only in app memory after querying all records. The matching response is Enforce tenant-scoped database authorization and key-based access, because safe multitenancy requires robust data-level boundaries.

The competing option Keep dedicated databases/accounts according to compliance rather than pooling would be useful when tenants require physical data separation by legal contract. Isolation requirements override cost optimization. The present issue is customers' usage swings unpredictably throughout the day; it is more directly addressed by Consider Aurora Serverless v2 or DynamoDB on-demand where workload semantics fit. Elastic pricing aligns capacity with variable demand.

Enforce tenant-scoped database authorization and key-based access would help if a shared table filters tenants only in app memory after querying all records. Safe multitenancy requires robust data-level boundaries. That is not the symptom described here: customers' usage swings unpredictably throughout the day. The matching response is Consider Aurora Serverless v2 or DynamoDB on-demand where workload semantics fit, because elastic pricing aligns capacity with variable demand.

For the alternative failure storage costs rise because tenant audit histories are retained indefinitely, the control Implement compliant tiering and expiration for historical records is relevant: data lifecycle policies control unbounded growth. Here the evidence instead is customers' usage swings unpredictably throughout the day. The selected technical mechanism is Consider Aurora Serverless v2 or DynamoDB on-demand where workload semantics fit: elastic pricing aligns capacity with variable demand.

Enforce tenant-scoped database authorization and key-based access would help if a shared table filters tenants only in app memory after querying all records. Safe multitenancy requires robust data-level boundaries. That is not the symptom described here: storage costs rise because tenant audit histories are retained indefinitely. The matching response is Implement compliant tiering and expiration for historical records, because data lifecycle policies control unbounded growth.

For the alternative failure customers' usage swings unpredictably throughout the day, the control Consider Aurora Serverless v2 or DynamoDB on-demand where workload semantics fit is relevant: elastic pricing aligns capacity with variable demand. Here the evidence instead is storage costs rise because tenant audit histories are retained indefinitely. The selected technical mechanism is Implement compliant tiering and expiration for historical records: data lifecycle policies control unbounded growth.

Keep dedicated databases/accounts according to compliance rather than pooling solves a different operational problem where tenants require physical data separation by legal contract. Isolation requirements override cost optimization. It does not directly address storage costs rise because tenant audit histories are retained indefinitely. The case calls for Implement compliant tiering and expiration for historical records, since data lifecycle policies control unbounded growth.