Amazon AWS Certified Solutions Architect Associate SAA-C03 Database Capacity Retention and Migration Economics Practice Test

 

Topic 18 covers database capacity retention and migration economics for the AWS Certified Solutions Architect – Associate certification. These original practice questions apply the verified SAA-C03 objectives to practical decisions and troubleshooting. Select one answer unless a fixed number is requested. For broader preparation, visit the AWS Certified Solutions Architect Associate SAA-C03 Exam Dumps page. Each option includes an explanation of the relevant behavior and scenario constraints.

Question 1

A new DynamoDB table has unpredictable traffic from zero to large spikes and the team cannot forecast capacity. Which mode is the simpler economic fit?

1. Use DynamoDB on-demand capacity mode.
2. Use DynamoDB provisioned capacity with auto scaling when the predictable pattern makes that model economical.
3. Evaluate an Aurora/serverless scaling option that matches the chosen engine and workload constraints.
4. Right-size the DB compute class based on measured CPU requirements and required headroom.
5. Stop and start the nonproduction RDS instance on a schedule where the engine and operational requirements support it.

Correct Answer: A

 

Correct Answer

Answer A is correct because On-demand charges per request and automatically accommodates changing throughput without pre-provisioning capacity. This directly meets the decisive requirement: DynamoDB on-demand.

Incorrect Answers

Answer B is incorrect because Provisioned mode bills configured capacity and can use auto scaling; predictable sustained utilization is where that model can be advantageous. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: DynamoDB on-demand.

Answer C is incorrect because Elastic database capacity can avoid continuously paying for a peak-sized relational fleet when traffic varies dramatically. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: DynamoDB on-demand.

Answer D is incorrect because The bottleneck is compute, so changing unrelated storage capacity does not address performance or optimize the right cost dimension. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: DynamoDB on-demand.

Answer E is incorrect because Removing unneeded running hours can reduce compute cost while storage continues to persist. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: DynamoDB on-demand.

 

Question 2

A DynamoDB table has a stable high request rate every hour and the team can tune capacity with auto scaling. Which mode should be evaluated for predictable utilization economics?

1. Use DynamoDB on-demand capacity mode.
2. Use DynamoDB provisioned capacity with auto scaling when the predictable pattern makes that model economical.
3. Stop and start the nonproduction RDS instance on a schedule where the engine and operational requirements support it.
4. Right-size the DB compute class based on measured CPU requirements and required headroom.
5. Evaluate an Aurora/serverless scaling option that matches the chosen engine and workload constraints.

Correct Answer: B

 

Correct Answer

Answer B is correct because Provisioned mode bills configured capacity and can use auto scaling; predictable sustained utilization is where that model can be advantageous. This directly meets the decisive requirement: provisioned capacity.

Incorrect Answers

Answer A is incorrect because On-demand charges per request and automatically accommodates changing throughput without pre-provisioning capacity. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: provisioned capacity.

Answer C is incorrect because Removing unneeded running hours can reduce compute cost while storage continues to persist. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: provisioned capacity.

Answer D is incorrect because The bottleneck is compute, so changing unrelated storage capacity does not address performance or optimize the right cost dimension. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: provisioned capacity.

Answer E is incorrect because Elastic database capacity can avoid continuously paying for a peak-sized relational fleet when traffic varies dramatically. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: provisioned capacity.

 

Question 3

An RDS development database is used only 8 hours each weekday and can be stopped outside that window within service limitations. Which action can reduce idle instance hours?

1. Right-size the DB compute class based on measured CPU requirements and required headroom.
2. Evaluate an Aurora/serverless scaling option that matches the chosen engine and workload constraints.
3. Stop and start the nonproduction RDS instance on a schedule where the engine and operational requirements support it.
4. Use DynamoDB on-demand capacity mode.
5. Use DynamoDB provisioned capacity with auto scaling when the predictable pattern makes that model economical.

Correct Answer: C

 

Correct Answer

Answer C is correct because Removing unneeded running hours can reduce compute cost while storage continues to persist. This directly meets the decisive requirement: schedule stop/start where supported.

Incorrect Answers

Answer A is incorrect because The bottleneck is compute, so changing unrelated storage capacity does not address performance or optimize the right cost dimension. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: schedule stop/start where supported.

Answer B is incorrect because Elastic database capacity can avoid continuously paying for a peak-sized relational fleet when traffic varies dramatically. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: schedule stop/start where supported.

Answer D is incorrect because On-demand charges per request and automatically accommodates changing throughput without pre-provisioning capacity. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: schedule stop/start where supported.

Answer E is incorrect because Provisioned mode bills configured capacity and can use auto scaling; predictable sustained utilization is where that model can be advantageous. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: schedule stop/start where supported.

 

Question 4

A seasonal application needs relational SQL but sits nearly idle for months and then scales sharply. Which architecture should be evaluated rather than permanently sizing for peak?

1. Use DynamoDB on-demand capacity mode.
2. Evaluate an Aurora/serverless scaling option that matches the chosen engine and workload constraints.
3. Stop and start the nonproduction RDS instance on a schedule where the engine and operational requirements support it.
4. Right-size the DB compute class based on measured CPU requirements and required headroom.
5. Use DynamoDB provisioned capacity with auto scaling when the predictable pattern makes that model economical.

Correct Answer: B

 

Correct Answer

Answer B is correct because Elastic database capacity can avoid continuously paying for a peak-sized relational fleet when traffic varies dramatically. This directly meets the decisive requirement: Aurora/serverless scaling option if engine/requirements fit.

Incorrect Answers

Answer A is incorrect because On-demand charges per request and automatically accommodates changing throughput without pre-provisioning capacity. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Aurora/serverless scaling option if engine/requirements fit.

Answer C is incorrect because Removing unneeded running hours can reduce compute cost while storage continues to persist. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Aurora/serverless scaling option if engine/requirements fit.

Answer D is incorrect because The bottleneck is compute, so changing unrelated storage capacity does not address performance or optimize the right cost dimension. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Aurora/serverless scaling option if engine/requirements fit.

Answer E is incorrect because Provisioned mode bills configured capacity and can use auto scaling; predictable sustained utilization is where that model can be advantageous. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Aurora/serverless scaling option if engine/requirements fit.

 

Question 5

A production database is consistently CPU-bound at 80–90% while storage and I/O have headroom. Which cost action is appropriate?

1. Use DynamoDB on-demand capacity mode.
2. Right-size the DB compute class based on measured CPU requirements and required headroom.
3. Evaluate an Aurora/serverless scaling option that matches the chosen engine and workload constraints.
4. Stop and start the nonproduction RDS instance on a schedule where the engine and operational requirements support it.
5. Use DynamoDB provisioned capacity with auto scaling when the predictable pattern makes that model economical.

Correct Answer: B

 

Correct Answer

Answer B is correct because The bottleneck is compute, so changing unrelated storage capacity does not address performance or optimize the right cost dimension. This directly meets the decisive requirement: right-size to suitable DB instance class rather than overprovision storage.

Incorrect Answers

Answer A is incorrect because On-demand charges per request and automatically accommodates changing throughput without pre-provisioning capacity. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: right-size to suitable DB instance class rather than overprovision storage.

Answer C is incorrect because Elastic database capacity can avoid continuously paying for a peak-sized relational fleet when traffic varies dramatically. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: right-size to suitable DB instance class rather than overprovision storage.

Answer D is incorrect because Removing unneeded running hours can reduce compute cost while storage continues to persist. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: right-size to suitable DB instance class rather than overprovision storage.

Answer E is incorrect because Provisioned mode bills configured capacity and can use auto scaling; predictable sustained utilization is where that model can be advantageous. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: right-size to suitable DB instance class rather than overprovision storage.

 

Question 6

A product catalog receives millions of identical reads for the same small set of values and updates are infrequent. Which layer can reduce repeated database reads?

1. Use Amazon ElastiCache for the frequently repeated read data when consistency requirements allow it.
2. Replace full-table scans with key-based DynamoDB operations when the access pattern has the key.
3. Remove the unused replica after confirming it is not required for reads, failover, or recovery.
4. Return only the attributes and rows the application actually uses.
5. Use Amazon RDS Proxy.

Correct Answer: A

 

Correct Answer

Answer A is correct because A cache serves hot values without executing the same database read for every request. This directly meets the decisive requirement: ElastiCache.

Incorrect Answers

Answer B is incorrect because Query/GetItem access avoids consuming capacity to examine unrelated table items. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: ElastiCache.

Answer C is incorrect because A replica that satisfies no technical objective is ongoing database and storage spend without business value. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: ElastiCache.

Answer D is incorrect because Reducing result size cuts database processing and data transfer without changing the storage engine. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: ElastiCache.

Answer E is incorrect because RDS Proxy manages a pool of database connections so large numbers of application clients can reuse fewer underlying connections. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: ElastiCache.

 

Question 7

Thousands of short-lived Lambda invocations each open a new RDS connection and connection storms exhaust the database. Which managed component can pool connections?

1. Return only the attributes and rows the application actually uses.
2. Use Amazon ElastiCache for the frequently repeated read data when consistency requirements allow it.
3. Replace full-table scans with key-based DynamoDB operations when the access pattern has the key.
4. Remove the unused replica after confirming it is not required for reads, failover, or recovery.
5. Use Amazon RDS Proxy.

Correct Answer: E

 

Correct Answer

Answer E is correct because RDS Proxy manages a pool of database connections so large numbers of application clients can reuse fewer underlying connections. This directly meets the decisive requirement: RDS Proxy.

Incorrect Answers

Answer A is incorrect because Reducing result size cuts database processing and data transfer without changing the storage engine. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: RDS Proxy.

Answer B is incorrect because A cache serves hot values without executing the same database read for every request. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: RDS Proxy.

Answer C is incorrect because Query/GetItem access avoids consuming capacity to examine unrelated table items. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: RDS Proxy.

Answer D is incorrect because A replica that satisfies no technical objective is ongoing database and storage spend without business value. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: RDS Proxy.

 

Question 8

An RDS read replica is 2% utilized for months and no reporting or failover requirement uses it. Which cost action is appropriate?

1. Use Amazon ElastiCache for the frequently repeated read data when consistency requirements allow it.
2. Remove the unused replica after confirming it is not required for reads, failover, or recovery.
3. Use Amazon RDS Proxy.
4. Return only the attributes and rows the application actually uses.
5. Replace full-table scans with key-based DynamoDB operations when the access pattern has the key.

Correct Answer: B

 

Correct Answer

Answer B is correct because A replica that satisfies no technical objective is ongoing database and storage spend without business value. This directly meets the decisive requirement: remove replica after validation.

Incorrect Answers

Answer A is incorrect because A cache serves hot values without executing the same database read for every request. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: remove replica after validation.

Answer C is incorrect because RDS Proxy manages a pool of database connections so large numbers of application clients can reuse fewer underlying connections. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: remove replica after validation.

Answer D is incorrect because Reducing result size cuts database processing and data transfer without changing the storage engine. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: remove replica after validation.

Answer E is incorrect because Query/GetItem access avoids consuming capacity to examine unrelated table items. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: remove replica after validation.

 

Question 9

An application repeatedly scans an entire DynamoDB table although every request has the partition key. Which change reduces unnecessary capacity consumption?

1. Remove the unused replica after confirming it is not required for reads, failover, or recovery.
2. Return only the attributes and rows the application actually uses.
3. Replace full-table scans with key-based DynamoDB operations when the access pattern has the key.
4. Use Amazon ElastiCache for the frequently repeated read data when consistency requirements allow it.
5. Use Amazon RDS Proxy.

Correct Answer: C

 

Correct Answer

Answer C is correct because Query/GetItem access avoids consuming capacity to examine unrelated table items. This directly meets the decisive requirement: use key-based Query/GetItem access.

Incorrect Answers

Answer A is incorrect because A replica that satisfies no technical objective is ongoing database and storage spend without business value. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: use key-based Query/GetItem access.

Answer B is incorrect because Reducing result size cuts database processing and data transfer without changing the storage engine. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: use key-based Query/GetItem access.

Answer D is incorrect because A cache serves hot values without executing the same database read for every request. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: use key-based Query/GetItem access.

Answer E is incorrect because RDS Proxy manages a pool of database connections so large numbers of application clients can reuse fewer underlying connections. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: use key-based Query/GetItem access.

 

Question 10

An Aurora workload retrieves 2 MB rows when each API call uses only three small columns. Which optimization can reduce database and network work?

1. Use Amazon ElastiCache for the frequently repeated read data when consistency requirements allow it.
2. Remove the unused replica after confirming it is not required for reads, failover, or recovery.
3. Use Amazon RDS Proxy.
4. Return only the attributes and rows the application actually uses.
5. Replace full-table scans with key-based DynamoDB operations when the access pattern has the key.

Correct Answer: D

 

Correct Answer

Answer D is correct because Reducing result size cuts database processing and data transfer without changing the storage engine. This directly meets the decisive requirement: query only needed attributes/rows.

Incorrect Answers

Answer A is incorrect because A cache serves hot values without executing the same database read for every request. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: query only needed attributes/rows.

Answer B is incorrect because A replica that satisfies no technical objective is ongoing database and storage spend without business value. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: query only needed attributes/rows.

Answer C is incorrect because RDS Proxy manages a pool of database connections so large numbers of application clients can reuse fewer underlying connections. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: query only needed attributes/rows.

Answer E is incorrect because Query/GetItem access avoids consuming capacity to examine unrelated table items. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: query only needed attributes/rows.

 

Question 11

A commercial database license is a major cost driver and the application is compatible with PostgreSQL after testing. Which migration direction can reduce license expense?

1. Use an analytical warehouse such as Amazon Redshift when the workload is primarily large-scale analytical SQL.
2. Use a managed relational service such as Amazon RDS or Aurora with a compatible engine.
3. Migrate to a managed open-source relational engine such as RDS for PostgreSQL when compatibility testing succeeds.
4. Evaluate Amazon Aurora.
5. Use Amazon DynamoDB.

Correct Answer: C

 

Correct Answer

Answer C is correct because Removing proprietary-license dependency can reduce total database cost while retaining relational behavior. This directly meets the decisive requirement: managed open-source PostgreSQL.

Incorrect Answers

Answer A is incorrect because Separating analytical work from an OLTP engine avoids paying to over-scale the wrong database architecture. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: managed open-source PostgreSQL.

Answer B is incorrect because Relational constraints and existing SQL semantics are strong requirements and should not be discarded solely for a different billing model. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: managed open-source PostgreSQL.

Answer D is incorrect because Aurora provides MySQL/PostgreSQL-compatible managed database clusters with distributed storage and read scaling features. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: managed open-source PostgreSQL.

Answer E is incorrect because DynamoDB is a managed key-value/document database designed for low-latency access and elastic throughput without relational joins. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: managed open-source PostgreSQL.

 

Question 12

A key-value session store needs single-digit millisecond access at unpredictable scale and does not require joins. Which managed database model best matches?

1. Evaluate Amazon Aurora.
2. Use a managed relational service such as Amazon RDS or Aurora with a compatible engine.
3. Use an analytical warehouse such as Amazon Redshift when the workload is primarily large-scale analytical SQL.
4. Use Amazon DynamoDB.
5. Migrate to a managed open-source relational engine such as RDS for PostgreSQL when compatibility testing succeeds.

Correct Answer: D

 

Correct Answer

Answer D is correct because DynamoDB is a managed key-value/document database designed for low-latency access and elastic throughput without relational joins. This directly meets the decisive requirement: DynamoDB.

Incorrect Answers

Answer A is incorrect because Aurora provides MySQL/PostgreSQL-compatible managed database clusters with distributed storage and read scaling features. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: DynamoDB.

Answer B is incorrect because Relational constraints and existing SQL semantics are strong requirements and should not be discarded solely for a different billing model. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: DynamoDB.

Answer C is incorrect because Separating analytical work from an OLTP engine avoids paying to over-scale the wrong database architecture. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: DynamoDB.

Answer E is incorrect because Removing proprietary-license dependency can reduce total database cost while retaining relational behavior. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: DynamoDB.

 

Question 13

A transactional application requires standard relational SQL, foreign keys, and an existing MySQL-compatible driver. Which managed family should be evaluated before adopting a NoSQL engine?

1. Use a managed relational service such as Amazon RDS or Aurora with a compatible engine.
2. Use an analytical warehouse such as Amazon Redshift when the workload is primarily large-scale analytical SQL.
3. Evaluate Amazon Aurora.
4. Use Amazon DynamoDB.
5. Migrate to a managed open-source relational engine such as RDS for PostgreSQL when compatibility testing succeeds.

Correct Answer: A

 

Correct Answer

Answer A is correct because Relational constraints and existing SQL semantics are strong requirements and should not be discarded solely for a different billing model. This directly meets the decisive requirement: RDS/Aurora relational engine.

Incorrect Answers

Answer B is incorrect because Separating analytical work from an OLTP engine avoids paying to over-scale the wrong database architecture. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: RDS/Aurora relational engine.

Answer C is incorrect because Aurora provides MySQL/PostgreSQL-compatible managed database clusters with distributed storage and read scaling features. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: RDS/Aurora relational engine.

Answer D is incorrect because DynamoDB is a managed key-value/document database designed for low-latency access and elastic throughput without relational joins. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: RDS/Aurora relational engine.

Answer E is incorrect because Removing proprietary-license dependency can reduce total database cost while retaining relational behavior. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: RDS/Aurora relational engine.

 

Question 14

A reporting workload performs large analytical scans and aggregations over historical data, with little OLTP activity. Which database category is more appropriate than scaling an OLTP instance indefinitely?

1. Use Amazon DynamoDB.
2. Evaluate Amazon Aurora.
3. Use a managed relational service such as Amazon RDS or Aurora with a compatible engine.
4. Migrate to a managed open-source relational engine such as RDS for PostgreSQL when compatibility testing succeeds.
5. Use an analytical warehouse such as Amazon Redshift when the workload is primarily large-scale analytical SQL.

Correct Answer: E

 

Correct Answer

Answer E is correct because Separating analytical work from an OLTP engine avoids paying to over-scale the wrong database architecture. This directly meets the decisive requirement: analytical warehouse such as Redshift.

Incorrect Answers

Answer A is incorrect because DynamoDB is a managed key-value/document database designed for low-latency access and elastic throughput without relational joins. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: analytical warehouse such as Redshift.

Answer B is incorrect because Aurora provides MySQL/PostgreSQL-compatible managed database clusters with distributed storage and read scaling features. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: analytical warehouse such as Redshift.

Answer C is incorrect because Relational constraints and existing SQL semantics are strong requirements and should not be discarded solely for a different billing model. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: analytical warehouse such as Redshift.

Answer D is incorrect because Removing proprietary-license dependency can reduce total database cost while retaining relational behavior. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: analytical warehouse such as Redshift.

 

Question 15

A team wants MySQL/PostgreSQL compatibility plus a cloud-native managed cluster that separates compute from distributed storage and can add read replicas. Which service is a candidate?

1. Use a managed relational service such as Amazon RDS or Aurora with a compatible engine.
2. Use Amazon DynamoDB.
3. Evaluate Amazon Aurora.
4. Migrate to a managed open-source relational engine such as RDS for PostgreSQL when compatibility testing succeeds.
5. Use an analytical warehouse such as Amazon Redshift when the workload is primarily large-scale analytical SQL.

Correct Answer: C

 

Correct Answer

Answer C is correct because Aurora provides MySQL/PostgreSQL-compatible managed database clusters with distributed storage and read scaling features. This directly meets the decisive requirement: Aurora.

Incorrect Answers

Answer A is incorrect because Relational constraints and existing SQL semantics are strong requirements and should not be discarded solely for a different billing model. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Aurora.

Answer B is incorrect because DynamoDB is a managed key-value/document database designed for low-latency access and elastic throughput without relational joins. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Aurora.

Answer D is incorrect because Removing proprietary-license dependency can reduce total database cost while retaining relational behavior. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Aurora.

Answer E is incorrect because Separating analytical work from an OLTP engine avoids paying to over-scale the wrong database architecture. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Aurora.

 

Question 16

Automated RDS backups are retained for the maximum period although policy requires only seven days. Which setting should be reviewed?

1. Set automated backup retention to the actual recovery and policy requirement.
2. Implement ownership and retention cleanup for manual snapshots from retired databases.
3. Review the retained snapshot inventory and enforce a backup lifecycle.
4. Keep frequent recovery points only for the short operational window and retain the required monthly long-term snapshots separately.
5. Remove the replica if it serves no read, availability, or recovery requirement beyond already-sufficient backups.

Correct Answer: A

 

Correct Answer

Answer A is correct because Keeping automated backups longer than required consumes backup storage without adding approved recovery value. This directly meets the decisive requirement: reduce retention to approved requirement.

Incorrect Answers

Answer B is incorrect because Manual snapshots persist until explicitly removed and can become orphaned cost when decommissioning processes ignore them. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: reduce retention to approved requirement.

Answer C is incorrect because Unbounded snapshot retention is a direct source of backup-storage growth independent of the live database size. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: reduce retention to approved requirement.

Answer D is incorrect because Tiered retention aligns storage cost with the different value of recent and compliance recovery points. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: reduce retention to approved requirement.

Answer E is incorrect because A continuously running replica is generally more expensive than backup storage when no live replica capability is needed. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: reduce retention to approved requirement.

 

Question 17

Manual snapshots from decommissioned databases accumulate indefinitely. What process should be added?

1. Keep frequent recovery points only for the short operational window and retain the required monthly long-term snapshots separately.
2. Review the retained snapshot inventory and enforce a backup lifecycle.
3. Remove the replica if it serves no read, availability, or recovery requirement beyond already-sufficient backups.
4. Implement ownership and retention cleanup for manual snapshots from retired databases.
5. Set automated backup retention to the actual recovery and policy requirement.

Correct Answer: D

 

Correct Answer

Answer D is correct because Manual snapshots persist until explicitly removed and can become orphaned cost when decommissioning processes ignore them. This directly meets the decisive requirement: snapshot ownership/retention cleanup.

Incorrect Answers

Answer A is incorrect because Tiered retention aligns storage cost with the different value of recent and compliance recovery points. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: snapshot ownership/retention cleanup.

Answer B is incorrect because Unbounded snapshot retention is a direct source of backup-storage growth independent of the live database size. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: snapshot ownership/retention cleanup.

Answer C is incorrect because A continuously running replica is generally more expensive than backup storage when no live replica capability is needed. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: snapshot ownership/retention cleanup.

Answer E is incorrect because Keeping automated backups longer than required consumes backup storage without adding approved recovery value. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: snapshot ownership/retention cleanup.

 

Question 18

A compliance database needs monthly snapshots for seven years but hourly snapshots only for 30 days. Which policy is most economical?

1. Implement ownership and retention cleanup for manual snapshots from retired databases.
2. Set automated backup retention to the actual recovery and policy requirement.
3. Keep frequent recovery points only for the short operational window and retain the required monthly long-term snapshots separately.
4. Review the retained snapshot inventory and enforce a backup lifecycle.
5. Remove the replica if it serves no read, availability, or recovery requirement beyond already-sufficient backups.

Correct Answer: C

 

Correct Answer

Answer C is correct because Tiered retention aligns storage cost with the different value of recent and compliance recovery points. This directly meets the decisive requirement: tiered snapshot retention.

Incorrect Answers

Answer A is incorrect because Manual snapshots persist until explicitly removed and can become orphaned cost when decommissioning processes ignore them. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: tiered snapshot retention.

Answer B is incorrect because Keeping automated backups longer than required consumes backup storage without adding approved recovery value. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: tiered snapshot retention.

Answer D is incorrect because Unbounded snapshot retention is a direct source of backup-storage growth independent of the live database size. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: tiered snapshot retention.

Answer E is incorrect because A continuously running replica is generally more expensive than backup storage when no live replica capability is needed. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: tiered snapshot retention.

 

Question 19

A read replica is kept solely as a backup even though the recovery requirement is already met by backups and the replica is never queried. Which statement is correct?

1. Review the retained snapshot inventory and enforce a backup lifecycle.
2. Set automated backup retention to the actual recovery and policy requirement.
3. Keep frequent recovery points only for the short operational window and retain the required monthly long-term snapshots separately.
4. Implement ownership and retention cleanup for manual snapshots from retired databases.
5. Remove the replica if it serves no read, availability, or recovery requirement beyond already-sufficient backups.

Correct Answer: E

 

Correct Answer

Answer E is correct because A continuously running replica is generally more expensive than backup storage when no live replica capability is needed. This directly meets the decisive requirement: remove replica if no availability/read objective.

Incorrect Answers

Answer A is incorrect because Unbounded snapshot retention is a direct source of backup-storage growth independent of the live database size. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: remove replica if no availability/read objective.

Answer B is incorrect because Keeping automated backups longer than required consumes backup storage without adding approved recovery value. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: remove replica if no availability/read objective.

Answer C is incorrect because Tiered retention aligns storage cost with the different value of recent and compliance recovery points. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: remove replica if no availability/read objective.

Answer D is incorrect because Manual snapshots persist until explicitly removed and can become orphaned cost when decommissioning processes ignore them. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: remove replica if no availability/read objective.

 

Question 20

Backup storage grows faster than database data because old snapshots are never deleted. Which metric/process should be investigated?

1. Set automated backup retention to the actual recovery and policy requirement.
2. Keep frequent recovery points only for the short operational window and retain the required monthly long-term snapshots separately.
3. Review the retained snapshot inventory and enforce a backup lifecycle.
4. Remove the replica if it serves no read, availability, or recovery requirement beyond already-sufficient backups.
5. Implement ownership and retention cleanup for manual snapshots from retired databases.

Correct Answer: C

 

Correct Answer

Answer C is correct because Unbounded snapshot retention is a direct source of backup-storage growth independent of the live database size. This directly meets the decisive requirement: snapshot lifecycle and retained backup inventory.

Incorrect Answers

Answer A is incorrect because Keeping automated backups longer than required consumes backup storage without adding approved recovery value. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: snapshot lifecycle and retained backup inventory.

Answer B is incorrect because Tiered retention aligns storage cost with the different value of recent and compliance recovery points. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: snapshot lifecycle and retained backup inventory.

Answer D is incorrect because A continuously running replica is generally more expensive than backup storage when no live replica capability is needed. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: snapshot lifecycle and retained backup inventory.

Answer E is incorrect because Manual snapshots persist until explicitly removed and can become orphaned cost when decommissioning processes ignore them. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: snapshot lifecycle and retained backup inventory.

 

Question 21

A PostgreSQL database is moving from on premises to RDS PostgreSQL with only minutes of cutover downtime allowed. Which AWS service can perform full load plus ongoing CDC?

1. Measure available path throughput and applicable inter-Region/data-transfer economics before setting the migration method and window.
2. Plan schema, SQL, and application-code conversion in addition to copying rows.
3. Use AWS Database Migration Service with full load and change data capture.
4. Stop and delete migration resources that no longer serve a validation or production migration.
5. Use continuous replication/CDC after the initial load, then cut over after replication lag is acceptable.

Correct Answer: C

 

Correct Answer

Answer C is correct because DMS can load existing data and then replicate ongoing changes so the final cutover can occur with a smaller downtime window. This directly meets the decisive requirement: AWS DMS full load plus CDC.

Incorrect Answers

Answer A is incorrect because Migration duration and network cost depend on how much data crosses the path and how quickly the path can move it. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: AWS DMS full load plus CDC.

Answer B is incorrect because Changing database engines can require semantic changes that a transport service alone cannot resolve. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: AWS DMS full load plus CDC.

Answer D is incorrect because Replication infrastructure that is left running after the migration provides no ongoing value and continues to incur charges. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: AWS DMS full load plus CDC.

Answer E is incorrect because Keeping the target synchronized avoids an outage equal to the full bulk-transfer duration. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: AWS DMS full load plus CDC.

 

Question 22

A migration changes from a proprietary relational engine to PostgreSQL and many stored procedures are engine-specific. What work must be planned in addition to data movement?

1. Plan schema, SQL, and application-code conversion in addition to copying rows.
2. Measure available path throughput and applicable inter-Region/data-transfer economics before setting the migration method and window.
3. Stop and delete migration resources that no longer serve a validation or production migration.
4. Use AWS Database Migration Service with full load and change data capture.
5. Use continuous replication/CDC after the initial load, then cut over after replication lag is acceptable.

Correct Answer: A

 

Correct Answer

Answer A is correct because Changing database engines can require semantic changes that a transport service alone cannot resolve. This directly meets the decisive requirement: schema/application conversion assessment.

Incorrect Answers

Answer B is incorrect because Migration duration and network cost depend on how much data crosses the path and how quickly the path can move it. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: schema/application conversion assessment.

Answer C is incorrect because Replication infrastructure that is left running after the migration provides no ongoing value and continues to incur charges. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: schema/application conversion assessment.

Answer D is incorrect because DMS can load existing data and then replicate ongoing changes so the final cutover can occur with a smaller downtime window. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: schema/application conversion assessment.

Answer E is incorrect because Keeping the target synchronized avoids an outage equal to the full bulk-transfer duration. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: schema/application conversion assessment.

 

Question 23

A 20-TB database has a one-week migration window but the source changes continuously. Which approach minimizes a final outage compared with a one-time export/import?

1. Use continuous replication/CDC after the initial load, then cut over after replication lag is acceptable.
2. Plan schema, SQL, and application-code conversion in addition to copying rows.
3. Use AWS Database Migration Service with full load and change data capture.
4. Stop and delete migration resources that no longer serve a validation or production migration.
5. Measure available path throughput and applicable inter-Region/data-transfer economics before setting the migration method and window.

Correct Answer: A

 

Correct Answer

Answer A is correct because Keeping the target synchronized avoids an outage equal to the full bulk-transfer duration. This directly meets the decisive requirement: continuous replication then cutover.

Incorrect Answers

Answer B is incorrect because Changing database engines can require semantic changes that a transport service alone cannot resolve. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: continuous replication then cutover.

Answer C is incorrect because DMS can load existing data and then replicate ongoing changes so the final cutover can occur with a smaller downtime window. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: continuous replication then cutover.

Answer D is incorrect because Replication infrastructure that is left running after the migration provides no ongoing value and continues to incur charges. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: continuous replication then cutover.

Answer E is incorrect because Migration duration and network cost depend on how much data crosses the path and how quickly the path can move it. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: continuous replication then cutover.

 

Question 24

The target database is in another Region but source and target can communicate privately through existing network connectivity. Which cost/performance factor should be measured before choosing migration duration?

1. Use continuous replication/CDC after the initial load, then cut over after replication lag is acceptable.
2. Stop and delete migration resources that no longer serve a validation or production migration.
3. Use AWS Database Migration Service with full load and change data capture.
4. Plan schema, SQL, and application-code conversion in addition to copying rows.
5. Measure available path throughput and applicable inter-Region/data-transfer economics before setting the migration method and window.

Correct Answer: E

 

Correct Answer

Answer E is correct because Migration duration and network cost depend on how much data crosses the path and how quickly the path can move it. This directly meets the decisive requirement: network throughput and transfer cost.

Incorrect Answers

Answer A is incorrect because Keeping the target synchronized avoids an outage equal to the full bulk-transfer duration. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: network throughput and transfer cost.

Answer B is incorrect because Replication infrastructure that is left running after the migration provides no ongoing value and continues to incur charges. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: network throughput and transfer cost.

Answer C is incorrect because DMS can load existing data and then replicate ongoing changes so the final cutover can occur with a smaller downtime window. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: network throughput and transfer cost.

Answer D is incorrect because Changing database engines can require semantic changes that a transport service alone cannot resolve. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: network throughput and transfer cost.

 

Question 25

A test migration keeps the DMS replication resources running for months after validation completes. Which action should be taken?

1. Stop and delete migration resources that no longer serve a validation or production migration.
2. Plan schema, SQL, and application-code conversion in addition to copying rows.
3. Use continuous replication/CDC after the initial load, then cut over after replication lag is acceptable.
4. Measure available path throughput and applicable inter-Region/data-transfer economics before setting the migration method and window.
5. Use AWS Database Migration Service with full load and change data capture.

Correct Answer: A

 

Correct Answer

Answer A is correct because Replication infrastructure that is left running after the migration provides no ongoing value and continues to incur charges. This directly meets the decisive requirement: decommission unused replication resources.

Incorrect Answers

Answer B is incorrect because Changing database engines can require semantic changes that a transport service alone cannot resolve. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: decommission unused replication resources.

Answer C is incorrect because Keeping the target synchronized avoids an outage equal to the full bulk-transfer duration. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: decommission unused replication resources.

Answer D is incorrect because Migration duration and network cost depend on how much data crosses the path and how quickly the path can move it. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: decommission unused replication resources.

Answer E is incorrect because DMS can load existing data and then replicate ongoing changes so the final cutover can occur with a smaller downtime window. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: decommission unused replication resources.

 

Question 26

Two RDS designs have similar instance prices, but one requires twice the provisioned IOPS and much more backup storage. Which comparison is correct?

1. Use Cost Explorer or detailed CUR data to separate database, storage, I/O, backup, transfer, and related service charges.
2. On-demand DynamoDB cost rises with the consumed read/write request units for the increased workload.
3. Include RDS Proxy pricing together with the database connection-efficiency and operational benefits it provides.
4. Compare instance, storage, provisioned I/O/throughput, backup, data-transfer, and dependent-service cost rather than instance price alone.
5. Include inter-Region replication/data-transfer charges.

Correct Answer: D

 

Correct Answer

Answer D is correct because Database total cost is composed of multiple charge dimensions that can outweigh a small compute-price difference. This directly meets the decisive requirement: compare total database stack cost.

Incorrect Answers

Answer A is incorrect because Cost-management data lets the team attribute the increase instead of optimizing the wrong component. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: compare total database stack cost.

Answer B is incorrect because PAY_PER_REQUEST billing ties throughput cost to requests rather than to provisioned capacity held idle. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: compare total database stack cost.

Answer C is incorrect because Adding a managed proxy creates its own charge and should be justified by the workload problem it solves. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: compare total database stack cost.

Answer E is incorrect because Global replicas move data between Regions, so network replication cost is part of the total design economics. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: compare total database stack cost.

 

Question 27

A DynamoDB table uses on-demand mode and traffic triples. Which cost driver changes directly with the new workload?

1. Include inter-Region replication/data-transfer charges.
2. Compare instance, storage, provisioned I/O/throughput, backup, data-transfer, and dependent-service cost rather than instance price alone.
3. Use Cost Explorer or detailed CUR data to separate database, storage, I/O, backup, transfer, and related service charges.
4. Include RDS Proxy pricing together with the database connection-efficiency and operational benefits it provides.
5. On-demand DynamoDB cost rises with the consumed read/write request units for the increased workload.

Correct Answer: E

 

Correct Answer

Answer E is correct because PAY_PER_REQUEST billing ties throughput cost to requests rather than to provisioned capacity held idle. This directly meets the decisive requirement: request consumption.

Incorrect Answers

Answer A is incorrect because Global replicas move data between Regions, so network replication cost is part of the total design economics. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: request consumption.

Answer B is incorrect because Database total cost is composed of multiple charge dimensions that can outweigh a small compute-price difference. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: request consumption.

Answer C is incorrect because Cost-management data lets the team attribute the increase instead of optimizing the wrong component. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: request consumption.

Answer D is incorrect because Adding a managed proxy creates its own charge and should be justified by the workload problem it solves. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: request consumption.

 

Question 28

A proposed RDS Proxy solves connection pressure but the team assumes it is free. What should the cost review include?

1. Compare instance, storage, provisioned I/O/throughput, backup, data-transfer, and dependent-service cost rather than instance price alone.
2. Include inter-Region replication/data-transfer charges.
3. Include RDS Proxy pricing together with the database connection-efficiency and operational benefits it provides.
4. Use Cost Explorer or detailed CUR data to separate database, storage, I/O, backup, transfer, and related service charges.
5. On-demand DynamoDB cost rises with the consumed read/write request units for the increased workload.

Correct Answer: C

 

Correct Answer

Answer C is correct because Adding a managed proxy creates its own charge and should be justified by the workload problem it solves. This directly meets the decisive requirement: proxy charges plus database savings/benefit.

Incorrect Answers

Answer A is incorrect because Database total cost is composed of multiple charge dimensions that can outweigh a small compute-price difference. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: proxy charges plus database savings/benefit.

Answer B is incorrect because Global replicas move data between Regions, so network replication cost is part of the total design economics. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: proxy charges plus database savings/benefit.

Answer D is incorrect because Cost-management data lets the team attribute the increase instead of optimizing the wrong component. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: proxy charges plus database savings/benefit.

Answer E is incorrect because PAY_PER_REQUEST billing ties throughput cost to requests rather than to provisioned capacity held idle. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: proxy charges plus database savings/benefit.

 

Question 29

A cross-Region read replica supports global readers. Which non-instance cost must be considered in the architecture?

1. Compare instance, storage, provisioned I/O/throughput, backup, data-transfer, and dependent-service cost rather than instance price alone.
2. Use Cost Explorer or detailed CUR data to separate database, storage, I/O, backup, transfer, and related service charges.
3. On-demand DynamoDB cost rises with the consumed read/write request units for the increased workload.
4. Include inter-Region replication/data-transfer charges.
5. Include RDS Proxy pricing together with the database connection-efficiency and operational benefits it provides.

Correct Answer: D

 

Correct Answer

Answer D is correct because Global replicas move data between Regions, so network replication cost is part of the total design economics. This directly meets the decisive requirement: replication/data-transfer charges.

Incorrect Answers

Answer A is incorrect because Database total cost is composed of multiple charge dimensions that can outweigh a small compute-price difference. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: replication/data-transfer charges.

Answer B is incorrect because Cost-management data lets the team attribute the increase instead of optimizing the wrong component. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: replication/data-transfer charges.

Answer C is incorrect because PAY_PER_REQUEST billing ties throughput cost to requests rather than to provisioned capacity held idle. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: replication/data-transfer charges.

Answer E is incorrect because Adding a managed proxy creates its own charge and should be justified by the workload problem it solves. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: replication/data-transfer charges.

 

Question 30

An Aurora cluster has many read replicas, high backup retention, and substantial I/O. Which tool can help identify which cost categories increased?

1. Include RDS Proxy pricing together with the database connection-efficiency and operational benefits it provides.
2. Use Cost Explorer or detailed CUR data to separate database, storage, I/O, backup, transfer, and related service charges.
3. Include inter-Region replication/data-transfer charges.
4. Compare instance, storage, provisioned I/O/throughput, backup, data-transfer, and dependent-service cost rather than instance price alone.
5. On-demand DynamoDB cost rises with the consumed read/write request units for the increased workload.

Correct Answer: B

 

Correct Answer

Answer B is correct because Cost-management data lets the team attribute the increase instead of optimizing the wrong component. This directly meets the decisive requirement: Cost Explorer/CUR analysis.

Incorrect Answers

Answer A is incorrect because Adding a managed proxy creates its own charge and should be justified by the workload problem it solves. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Cost Explorer/CUR analysis.

Answer C is incorrect because Global replicas move data between Regions, so network replication cost is part of the total design economics. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Cost Explorer/CUR analysis.

Answer D is incorrect because Database total cost is composed of multiple charge dimensions that can outweigh a small compute-price difference. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Cost Explorer/CUR analysis.

Answer E is incorrect because PAY_PER_REQUEST billing ties throughput cost to requests rather than to provisioned capacity held idle. It can be useful in a different cost or architecture situation, but it does not meet the decisive requirement here: Cost Explorer/CUR analysis.