DBS-C01 Practice Exam 59 — 20 Questions

Question 1

A company is deploying a new Amazon RDS for PostgreSQL DB instance. Which THREE actions are recommended for a secure deployment?

Accepted Answer

Enable encryption at rest using AWS KMS. Enabling encryption at rest using AWS KMS ensures that the underlying storage for the RDS for PostgreSQL DB instance is encrypted using AES-256 encryption. This protects data at rest against unauthorized physical access to the storage media and is a fundamental security best practice. KMS integration also allows for centralized key management and audit trails via AWS CloudTrail.

Answer

Use the default VPC for simplicity

Answer

Delete automated backups after the initial snapshot to reduce costs

Question 2

A media company is deploying a new video-on-demand catalog service on AWS. The catalog consists of millions of JSON documents, each with a unique video ID and nested metadata. The service requires single-digit millisecond read latency and must support a query pattern that filters videos by genre and release year. The operations team prefers a fully managed database that automatically scales storage and requires minimal administration. Which AWS database solution best meets these requirements?

Accepted Answer

Amazon DynamoDB with a primary key of video ID and a global secondary index on genre and release year.. The workload requires a fully managed, schema-flexible database that can store JSON documents, scale storage automatically, and deliver low-latency reads with filtering on non-key attributes. DynamoDB meets these needs with its document model and global secondary indexes, which enable efficient queries on genre and release year without managing servers or storage.

Answer

Amazon Aurora MySQL with a multi-AZ deployment and an Amazon ElastiCache for Redis cluster for caching.

Answer

Amazon Neptune with a graph data model and a Gremlin query endpoint for genre and year filtering.

Answer

Amazon RDS for PostgreSQL with the JSONB data type and a read replica for scaling reads.

Question 3

A healthcare analytics company ingests continuous patient vitals from thousands of bedside monitors. The data arrives as time-stamped measurements and must be queried by device ID and time range for real-time dashboards and long-term trend analysis. The team wants a fully managed AWS database that automatically partitions data by time and provides built-in retention policies. Which AWS database service should they choose?

Accepted Answer

Amazon Timestream. Amazon Timestream is the only fully managed time-series database among the choices. It automatically partitions data by time, enforces retention policies across memory and magnetic stores, and provides built-in functions for time-range queries. This directly matches the requirement for continuous vitals ingestion with real-time dashboards and long-term trend analysis without manual partition management.

Answer

Amazon RDS for PostgreSQL with the TimescaleDB extension

Answer

Amazon ElastiCache for Redis with sorted sets

Answer

Amazon DynamoDB with a partition key of device ID and sort key of timestamp

Question 4

A company is deploying a new application that uses Amazon RDS for PostgreSQL. The database must be highly available and fault-tolerant. Which deployment option meets these requirements?

Accepted Answer

Deploy the RDS instance with Multi-AZ configuration. Amazon RDS Multi-AZ deployment automatically provisions and maintains a synchronous standby replica in a different Availability Zone, providing automatic failover in the event of an AZ outage or primary instance failure. This configuration meets the requirements for high availability and fault tolerance by ensuring the database remains accessible with minimal downtime, as the standby replica is kept in sync synchronously and DNS is updated to point to the standby upon failover.

Answer

Deploy the RDS instance in a single Availability Zone

Answer

Deploy the RDS instance across multiple AWS Regions

Answer

Deploy the RDS instance with multiple read replicas

Question 5

A retail company runs its inventory management system on Amazon RDS for PostgreSQL. The application performs frequent updates to inventory counts. The operations team notices that write latency increases significantly during peak sales hours. The database is a single db.r5.large instance with General Purpose SSD (gp2) storage. The CPU utilization is around 40% during peaks, but the write latency spikes. The team suspects a storage bottleneck. Which change would most effectively reduce write latency?

Accepted Answer

Change storage type to Provisioned IOPS (io2) with sufficient IOPS.. The write latency spikes with CPU at 40% indicate a storage I/O bottleneck, not compute. gp2 volumes have burst IOPS based on volume size, and during peak sales the burst credits may be exhausted, causing latency. Changing to Provisioned IOPS (io2) with sufficient IOPS provides consistent, high IOPS and low latency for write-heavy workloads. This directly addresses the storage bottleneck.

Answer

Enable Multi-AZ deployment.

Answer

Add a read replica to offload read traffic.

Answer

Scale up to a db.r5.xlarge instance.

Question 6

A company needs to ensure that all changes to an Amazon RDS DB instance's security group are logged for auditing purposes. Which AWS service should be enabled?

Accepted Answer

AWS CloudTrail. AWS CloudTrail is the correct service because it records API calls made to the AWS environment, including actions like AuthorizeSecurityGroupIngress and RevokeSecurityGroupEgress that modify security group rules. When enabled, CloudTrail logs all changes to Amazon RDS security groups as management events, providing an immutable audit trail that can be stored in Amazon S3 for compliance and auditing purposes.

Answer

Amazon RDS Event Subscriptions

Answer

AWS Config

Answer

VPC Flow Logs

Question 7

A company uses Amazon DynamoDB to store session data for a globally distributed web application. The table is configured with on-demand capacity mode and has a single global secondary index (GSI) to support queries by user ID. The application team reports that during peak hours, some queries on the GSI return incomplete results. The database specialist needs to ensure that the GSI provides complete results. What is the most likely cause of the incomplete results?

Accepted Answer

The GSI is eventually consistent, and the queries are reading data before the index is updated.. Global secondary indexes are always eventually consistent. When data is written to the base table, the index is updated asynchronously. Queries on the GSI may not immediately reflect the latest writes, leading to incomplete results if the application expects strong consistency. To avoid this, the application can query the base table for strongly consistent reads or use a design that tolerates eventual consistency.

Answer

The table's on-demand capacity mode does not support global secondary indexes.

Answer

The GSI does not have enough read capacity units provisioned.

Answer

The GSI is missing some attributes that are required for the query.

Question 8

A healthcare analytics company ingests streaming vitals from bedside monitors into Amazon Kinesis Data Streams. The data must be queryable with sub-second latency for the last 30 days, and the analytics team writes ad-hoc SQL over device ID, patient cohort, and timestamp. The query volume is unpredictable and bursts to thousands of concurrent SQL statements during shift changes. Which database design meets these requirements with the LEAST operational overhead?

Accepted Answer

Load the stream into Amazon Timestream with a memory store retention of 30 days and a magnetic store retention of 365 days, then query it with Timestream SQL.. A managed time-series engine that keeps the queryable window in its fast tier is the lowest-overhead fit: it absorbs streaming writes, retains 30 days in memory for sub-second reads, and scales reads elastically without capacity planning. Relational, object-store, and key-value alternatives each force manual tuning or add latency that breaks the sub-second requirement.

Answer

Store events in an Amazon DynamoDB table with device ID as partition key and timestamp as sort key, and use PartiQL for the ad-hoc analytical queries.

Answer

Persist each event as an Amazon S3 object partitioned by device and hour, and query with Amazon Athena using a Glue Data Catalog table.

Answer

Write the stream to an Amazon RDS for PostgreSQL Multi-AZ instance with a 30-day partition scheme, and run the ad-hoc SQL directly against the primary.

Question 9

A media company is building a video-on-demand catalog service. The catalog data is highly structured and accessed by a REST API that performs many small, read-heavy queries with a strict requirement for single-digit millisecond latency at any scale. The team wants a fully managed AWS database that supports key-value access patterns, automatic scaling, and no server management. Which AWS database service BEST meets these requirements?

Accepted Answer

Amazon DynamoDB. DynamoDB is designed for key-value workloads that require predictable single-digit millisecond latency at any scale, with fully managed operations and automatic scaling. The other services either lack the required latency profile, are optimized for analytics, or are caches rather than primary databases, making DynamoDB the best fit.

Answer

Amazon Redshift

Answer

Amazon ElastiCache for Redis

Answer

Amazon RDS for MySQL with Multi-AZ

Question 10

A company is designing a new Amazon Aurora PostgreSQL database to support a global application with users in North America, Europe, and Asia. The application requires low-latency reads in each region and must survive a regional outage with minimal data loss. The company wants to minimize operational overhead. Which TWO actions should a database specialist take to meet these requirements? (Choose two.)

Accepted Answer

Create an Aurora global database with the primary cluster in one Region and secondary clusters in the other Regions.. An Aurora global database replicates data across Regions with low latency and supports fast failover, providing low-latency reads and disaster recovery. Using the regional reader endpoint in each Region directs read traffic to local replicas. Together, these actions meet the requirements for global low-latency reads and minimal data loss with minimal operational overhead.

Answer

Enable Aurora Auto Scaling for reader instances in each Region.

Answer

Use Amazon RDS for PostgreSQL with cross-Region read replicas.

Answer

Implement a multi-master Aurora cluster across all Regions.

Question 11

A retail company is building a new product catalog service. The catalog contains millions of items, and each item has a fixed set of attributes such as SKU, price, and category. The access pattern is key-based lookups by SKU at very high throughput, and the team wants single-digit millisecond latency with minimal operational effort. Which database should the team choose?

Accepted Answer

Amazon DynamoDB. The catalog has a fixed attribute set, is accessed by key at very high throughput, and needs low latency with minimal operations. DynamoDB is purpose-built for exactly this pattern: a partition key of SKU enables direct lookups, and the fully managed service scales horizontally without server administration. Relational, analytical, and cache services either add operational burden or do not match the access pattern.

Answer

Amazon RDS for MySQL with a Multi-AZ deployment

Answer

Amazon ElastiCache for Redis with persistence enabled

Answer

Amazon Redshift with a distribution key on SKU

Question 12

A financial analytics team runs complex aggregation queries over a 12 TB Amazon Redshift cluster. Analysts frequently join a large fact table to several dimension tables and also need to query semi-structured clickstream data stored as JSON. The team wants to minimize data movement during joins and avoid manually loading JSON into relational columns. Which combination of Redshift features should the team use?

Accepted Answer

Choose DISTKEY and SORTKEY columns that match the join and filter patterns, and use the SUPER data type with PartiQL to query JSON.. Colocating join keys through DISTKEY removes inter-node data movement, and a SORTKEY on filtered columns enables block pruning. The SUPER data type with PartiQL handles JSON natively inside Redshift, so analysts avoid manual parsing or loading. Distribution-all on a large fact table and even distribution both create unnecessary cost or network traffic, and VARCHAR parsing loses the flexibility the workload needs.

Answer

Use a compound SORTKEY on the fact table and enable concurrency scaling to handle JSON queries.

Answer

Use DISTSTYLE ALL on the fact table and query JSON with Redshift Spectrum against Amazon S3.

Answer

Use DISTSTYLE EVEN for all tables and store JSON as VARCHAR columns parsed with regular expressions.

Question 13

A media analytics company ingests roughly 40,000 JSON clickstream events per second into Amazon DynamoDB. Each event has a unique eventId, a sessionId, a userId, and a deviceType. Queries must retrieve all events for a given session ordered by event timestamp, and the table must scale horizontally without hot partitions. Which partition key design best satisfies these requirements?

Accepted Answer

Use sessionId as the partition key and eventTimestamp as the sort key.. The access pattern requires retrieving all events for a session in timestamp order, so the partition key should be sessionId and the sort key should be eventTimestamp. This gives even distribution across partitions and efficient ordered Query operations. Keys built from userId or raw timestamps either create hot partitions or fail to support the ordered session query.

Answer

Use userId as the partition key and deviceType as the sort key.

Answer

Use sessionId as the partition key and eventId as the sort key.

Answer

Use eventTimestamp as the partition key and sessionId as the sort key.

Question 14

A database specialist is designing an Amazon DynamoDB table for a real-time inventory system that must handle sudden traffic spikes and maintain consistent single-digit-millisecond latency. The team wants to minimize operational overhead and avoid capacity planning. Which TWO design choices should the specialist make? (Choose two.)

Accepted Answer

Use on-demand capacity mode for the table.. On-demand capacity mode removes capacity planning and scales automatically during spikes, while a high-cardinality, evenly accessed partition key prevents hot partitions that would otherwise throttle requests. Together they let the table sustain single-digit-millisecond latency with minimal operational effort. Fixed provisioning reintroduces planning, and Streams or blanket indexing do not address capacity or latency.

Answer

Provision read and write capacity with a fixed number of capacity units.

Answer

Create a local secondary index on every attribute to speed up queries.

Answer

Enable DynamoDB Streams to capture item-level changes.

Question 15

A database specialist manages an Amazon Aurora MySQL cluster that serves a read-heavy reporting application. The team needs to add a new Aurora Replica to the cluster and wants the replica to be automatically promoted if the primary writer instance fails. The specialist must also minimize the time the new replica takes to become available. Which action should the specialist take?

Accepted Answer

Add the Aurora Replica by specifying the existing DB cluster identifier, and assign it a higher promotion tier than the other replicas.. Aurora replicas share the cluster's distributed storage volume, so a replica added to the existing cluster becomes usable quickly. Promotion tiers define the order in which Aurora promotes replicas when the writer fails; a higher tier makes the new replica the preferred failover target. Combining quick attachment with a high promotion tier satisfies both availability and automatic promotion requirements.

Answer

Enable Aurora Auto Scaling on the cluster and set the replica's promotion tier to the lowest available value.

Answer

Add the Aurora Replica to the cluster's custom endpoint and configure the writer endpoint to fail over to it automatically.

Answer

Create the Aurora Replica by restoring a DB cluster snapshot to a new cluster, then add it to the existing cluster.

Question 16

A SaaS company is designing a multi-tenant analytics database on Amazon Redshift. Each tenant's data is stored in the same cluster, and tenants run ad hoc queries against their own data. The company must ensure that a tenant cannot read another tenant's rows, and must minimize the performance impact of one tenant's heavy queries on others. Which TWO design choices should be made? (Choose two.)

Accepted Answer

Use Redshift workload management (WLM) with separate query queues per tenant. The two requirements are data isolation and performance fairness in a shared Redshift cluster. Row-level security enforces per-tenant row visibility using tenant-scoped roles, and workload management assigns each tenant to its own query queue so heavy queries cannot starve others. Together they deliver secure multi-tenancy without duplicating clusters or data, which the other options either fail to provide or achieve at prohibitive cost.

Answer

Create a separate Amazon Redshift cluster for every tenant and use Amazon Redshift Spectrum to join across clusters

Answer

Enable concurrency scaling and rely on it alone to isolate tenant workloads

Answer

Store each tenant's data in a separate Amazon S3 bucket and query it directly with Amazon Athena

Question 17

A retail company runs an Amazon RDS for PostgreSQL database that serves an order management application. The application performs many short read queries for product lookups and a smaller number of heavy write transactions during flash sales. The team wants to offload read traffic and scale read capacity with minimal application changes, while keeping writes on the primary. Which approach should the team use?

Accepted Answer

Create one or more RDS read replicas and direct read queries to the replica endpoint using a reader endpoint.. RDS read replicas provide additional read-only capacity through asynchronous replication, and the reader endpoint spreads connections across them. The application only needs to point read queries at that endpoint, which is a minimal change. Multi-AZ standbys are for failover only, RDS Proxy pools connections rather than distributing reads, and migrating to Aurora is unnecessary overhead for this requirement.

Answer

Configure RDS Proxy in front of the primary instance to distribute read queries across connections.

Answer

Migrate the database to Amazon Aurora Serverless v2 and enable the reader endpoint.

Answer

Enable Multi-AZ with a standby instance and send read queries to the standby endpoint.

Question 18

A retail company runs an Amazon Aurora MySQL cluster for its product catalog. The catalog is read-mostly during business hours but receives a nightly bulk price update of roughly 4 million rows. During the update, application reads show elevated latency and replica lag spikes to several minutes. The team wants to reduce read impact and replica lag without changing the application's read endpoints or schema. Which change should a database specialist recommend?

Accepted Answer

Batch the nightly price update into smaller transactions committed at regular intervals instead of one large transaction.. Replica lag in Aurora tracks how quickly the change stream can be applied, and one multi-million-row transaction must be applied atomically, creating a single long lag window. Splitting the load into bounded commits lets replicas stay current continuously, which lowers read latency and lag without altering endpoints or schema. The other choices either redistribute reads, speed the writer, or add cross-Region hops without changing apply behavior.

Answer

Enable Aurora Global Database and point the application reader endpoint at the secondary Region's cluster.

Answer

Convert the nightly update to use LOAD DATA FROM S3 with the same single transaction boundary to speed up ingestion.

Answer

Increase the number of Aurora Replicas from two to five and place them behind the existing reader endpoint.

Question 19

A healthcare analytics team needs to store time-series vital-sign data from thousands of patient monitoring devices. Data arrives continuously at high velocity, is queried by time range and device ID, and is retained for 90 days before being archived to Amazon S3. The team wants a managed time-series database that supports SQL-like queries and automatic data lifecycle management. Which AWS service should they choose?

Accepted Answer

Amazon Timestream with a memory store and magnetic store retention policy. The workload demands high-velocity time-series ingestion, time-range and device queries, SQL-like access, and automatic retention tiering. Amazon Timestream provides these capabilities natively, including memory and magnetic storage tiers with configurable retention. The other options either lack native time-series optimization, require custom lifecycle logic, or are designed for batch analytics rather than continuous monitoring.

Answer

Amazon DynamoDB with a partition key of device ID and sort key of timestamp

Answer

Amazon RDS for PostgreSQL with the TimescaleDB extension

Answer

Amazon S3 with AWS Glue Data Catalog and Amazon Athena for queries

Question 20

A financial services firm is migrating a 4 TB on-premises Oracle database to Amazon RDS for Oracle. The migration window is 6 hours, and the source database can be taken offline only during that window. The firm wants to use AWS Database Migration Service (AWS DMS) to minimize cutover time. Which combination of tasks should be performed to meet the requirement?

Accepted Answer

Create a full-load plus change data capture (CDC) DMS task, let the full load complete, allow CDC to apply changes, then stop the application and switch over during the window.. A full-load plus CDC task moves the bulk of the data while the source stays online, then continuously applies changes. Only a short final catch-up and cutover require the application to be stopped, keeping the offline window within 6 hours. Full-load-only would miss changes, and CDC-only requires a pre-seeded consistent target, which is impractical here.

Answer

Create a full-load-only DMS task, run it during the migration window, then stop the application and switch the connection string to RDS.

Answer

Create a CDC-only DMS task, seed the target with an RMAN backup restored to RDS, then start the task to apply changes.

Answer

Create a full-load plus CDC DMS task, but configure the target to use Amazon RDS for Oracle with a read replica to offload the apply process.