DBS-C01 Management and Operations Practice Test 11 — 15 Questions

Question 1

A database administrator notices that the free storage space on an Amazon RDS for PostgreSQL DB instance is decreasing rapidly. The instance has automated backups enabled with a retention period of 7 days. Which action should be taken to monitor storage usage and prevent running out of space?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Set up a CloudWatch alarm on the FreeStorageSpace metric to notify when space is low.. Setting a CloudWatch alarm on the FreeStorageSpace metric allows proactive monitoring and notification when storage is low. Option A is incorrect because lifecycle policies to move backups to S3 Glacier are not applicable to RDS automated backups; they are managed separately. Option C is incorrect because increasing storage is a reactive measure, not a monitoring action, and may cause downtime. Option D is incorrect because Performance Insights focuses on database performance, not storage tracking.

Answer

Create a lifecycle policy to automatically move old backup files to Amazon S3 Glacier.

Answer

Increase the allocated storage for the DB instance immediately.

Answer

Enable Performance Insights to track storage usage.

Question 2

A company has an Amazon Redshift cluster with a single node. The cluster is used for reporting. Recently, queries have become slow, and the cluster's disk space is 80% full. Which action should be taken to improve query performance and manage storage?

Accepted Answer

Resize the cluster to include additional compute nodes.. Resizing the cluster to add compute nodes distributes data and workload across more nodes, improving query performance and increasing storage capacity. Option B is wrong because enabling compression (ENCODE AUTO) can reduce storage space but does not directly address the performance bottleneck caused by high disk usage on a single node; it also is not a quick fix for existing data. Option C is wrong because changing distribution style to ALL replicates all data to every node, which actually increases storage consumption and may worsen performance on a single-node cluster. Option D is wrong because running VACUUM reclaims space from deleted rows, but the cluster is 80% full and VACUUM does not add new storage capacity; it may provide temporary relief but not a long-term solution for performance.

Answer

Enable compression on all columns using the ENCODE AUTO option.

Answer

Modify the table's distribution style to DISTSTYLE ALL for all tables.

Answer

Run the VACUUM command to reclaim space from deleted rows.

Question 3

A company uses Amazon DynamoDB with global tables for a multi-region application. The application writes to the table in us-east-1. A developer notices that updates made in us-east-1 are not appearing in the replica in eu-west-1 after several minutes. Which action should be taken to diagnose the issue?

Accepted Answer

Review the CloudWatch metrics for ReplicationLatency and PendingReplicationCount for the global table.. CloudWatch metrics ReplicationLatency (the time lag between when an item is written to one replica and when it appears in another) and PendingReplicationCount (the number of items waiting to be replicated) directly help diagnose replication delays in DynamoDB global tables. Option B is incorrect because writes should continue to the local region (us-east-1); using the eu-west-1 endpoint would not resolve replication latency and could cause consistency issues. Option C is incorrect because DynamoDB Streams is automatically enabled for global tables with NEW_AND_OLD_IMAGES (not KEYS_ONLY) to replicate updates; checking the stream specification is not a diagnostic step for delays. Option D is incorrect because TTL (Time to Live) is for automatic item expiration, not related to replication performance or delays.

Answer

Modify the application to use the eu-west-1 endpoint for writes.

Answer

Check that the table has an active DynamoDB Streams stream with StreamSpecification set to KEYS_ONLY.

Answer

Verify that the TTL attribute on the table is correctly configured.

Question 4

A company is migrating an on-premises Oracle database to Amazon Aurora PostgreSQL. The database has several large tables with frequent INSERT and UPDATE operations. Which TWO actions should be taken to optimize performance after migration?

Accepted Answer

Enable autovacuum and configure it to run more frequently on the large tables.. Options A and B are correct. Enabling autovacuum and tuning it to run more frequently on large tables is essential in PostgreSQL to reclaim storage and prevent transaction ID wraparound, which is critical after a migration with high DML. Using Aurora PostgreSQL integration with Amazon S3 for bulk data loading is efficient because it leverages parallel processing and avoids network overhead. Option C is incorrect because setting synchronous_commit to ON reduces performance for write-heavy workloads by waiting for disk writes; OFF or REMOTE_WRITE is recommended. Option D is incorrect because RDS Proxy helps with connection management but does not directly optimize DML performance for large tables. Option E is incorrect because while partitioning can help query performance, it is not a direct optimization for INSERT/UPDATE operations and may add complexity; Aurora's storage layer already handles large tables efficiently.

Answer

Set synchronous_commit to ON to ensure data durability.

Answer

Deploy an RDS Proxy in front of the Aurora cluster to reduce connection overhead.

Answer

Partition the large tables by date to improve query performance.

Question 5

A company uses Amazon DynamoDB for a high-traffic gaming leaderboard. The table has a partition key of 'game_id' and a sort key of 'score'. During a tournament, the application experiences throttling on a single partition. The application uses strongly consistent reads. Which TWO actions should be taken to resolve the throttling and maintain performance?

Accepted Answer

Redesign the partition key to include a random suffix to distribute writes across partitions.. The correct actions are A and D. Option A: Redesigning the partition key to include a random suffix (e.g., appending a shard number) distributes write traffic across multiple partitions, preventing a single hot partition from throttling. Option D: For a leaderboard, eventually consistent reads are sufficient for non-critical queries, reducing read capacity consumption and alleviating read-side throttling. Option B (DAX) only caches reads and does not help with write throttling. Option C (increasing RCU) addresses read capacity, not write distribution to fix the hot partition. Option E (GSI) creates a separate index but does not redistribute writes on the base table's primary partition.

Answer

Enable DynamoDB Accelerator (DAX) to cache the most frequently read items.

Answer

Increase the provisioned read capacity units (RCU) for the table.

Answer

Create a global secondary index (GSI) with a different partition key.

Question 6

A company runs an Amazon RDS for MySQL DB instance with Multi-AZ. The application team reports that a recent schema change caused a significant increase in replication lag. The instance has a read replica in a different AWS Region. Which THREE actions should be taken to reduce replication lag?

Accepted Answer

Identify and optimize any long-running transactions or DDL statements that may be blocking replication.. Options A, C, and D are correct actions to reduce replication lag. A: Optimizing long-running transactions or DDL statements helps prevent blocking on the primary that can cause lag. C: Monitoring the ReplicaLag metric in CloudWatch allows you to identify trends and set alarms for proactive management. D: Setting the transaction isolation level to READ COMMITTED reduces locking on the primary, which can decrease replication lag. Option B is not the best action: increasing the instance class may improve performance but does not directly address the root cause of lag from schema changes. Option E is incorrect because disabling binary logging would stop replication entirely, not reduce lag.

Answer

Increase the DB instance class for both the primary and read replica to improve performance.

Answer

Disable binary logging on the primary instance to reduce I/O overhead.

Question 7

Refer to the exhibit. An IAM policy is attached to a role used by an application running on Amazon EC2. The application needs to read items from a DynamoDB table named 'MyTable' and create manual snapshots of an RDS instance. What is the issue with the policy?

Accepted Answer

The policy does not grant any DynamoDB read actions (e.g., GetItem, Query, Scan) required for the application.. The policy only grants 'PutItem' and 'DeleteItem' actions on the DynamoDB table, but the application needs to read items (e.g., 'GetItem' or 'Scan'). The missing DynamoDB read actions are the primary issue. Option B is incorrect because 'DescribeDBInstances' on all resources is not inherently too permissive for this use case; the key issue is the missing read permissions. Option C is incorrect because using 'Resource': '*' for DynamoDB is acceptable for read actions if the table ARN is not specified; it is not a policy structure problem. Option D is incorrect because 'CreateDBSnapshot' on '*' is allowed, but the policy correctly scopes RDS actions to 'Resource': '*' which includes all DB instances; the critical flaw remains the missing DynamoDB read actions.

Answer

The policy allows 'DescribeDBInstances' on all resources, which is too permissive.

Answer

The policy uses 'Resource': '*' for DynamoDB, which should be the table ARN.

Answer

The policy grants 'CreateDBSnapshot' on all resources, which is insufficient; it should be scoped to the specific DB instance.

Question 8

Refer to the exhibit. An IAM policy statement allows creating manual snapshots for an RDS instance. A database administrator is unable to create a snapshot from the AWS Management Console. The error message indicates insufficient permissions. What is the likely cause?

Accepted Answer

The policy does not include necessary read actions (e.g., 'DescribeDBInstances', 'DescribeDBSnapshots') that the console uses.. The AWS Management Console often performs multiple API calls (such as DescribeDBInstances and DescribeDBSnapshots) to list and display resources before allowing actions. Even if the CreateDBSnapshot action is allowed, the console will fail if it cannot describe the instance or snapshots. Option A is incorrect because the condition key 'aws:RequestedRegion' is correctly spelled. Option C is incorrect because the resource ARN correctly identifies the RDS instance for snapshot creation. Option D is incorrect because StringEquals is appropriate for matching the region exactly; StringLike is not needed.

Answer

The condition key 'aws:RequestedRegion' is misspelled.

Answer

The resource ARN is incorrect; it should include the snapshot ARN.

Answer

The condition uses 'StringEquals' but should use 'StringLike' for region matching.

Question 9

Refer to the exhibit. A database administrator runs the AWS CLI command shown. The output is: ["available", false]. What does this output indicate about the DB instance?

Accepted Answer

The DB instance is available and is not configured for Multi-AZ.. The output shows the DBInstanceStatus is 'available' and MultiAZ is false. Option A is incorrect because MultiAZ false indicates it is not in the process of being modified to enable Multi-AZ. Option B is incorrect because MultiAZ is false, meaning it is not in a Multi-AZ deployment. Option C is incorrect because the status is 'available', not 'stopped'.

Answer

The DB instance is in the process of being modified to enable Multi-AZ.

Answer

The DB instance is in a Multi-AZ deployment and is available.

Answer

The DB instance is stopped and is not in a Multi-AZ configuration.

Question 10

A company runs a critical application on Amazon RDS for PostgreSQL with Multi-AZ. The database has a large table (over 500 GB) that is frequently updated. The operations team notices that the primary instance's CPU usage is consistently above 90%, and the replica lag between the primary and standby is increasing during peak hours. The application can tolerate a few seconds of downtime. The team needs to reduce CPU load and improve write performance without changing the application code. Which action should be taken?

Accepted Answer

Upgrade the DB instance class to a larger size with more vCPUs.. Upgrading to a larger DB instance class provides additional vCPUs, reducing CPU utilization and alleviating replication lag. Option B is wrong: increasing the backup retention period does not reduce I/O during backups; it only keeps backups longer and does not address CPU or write performance. Option C is wrong: Multi-AZ RDS for PostgreSQL uses synchronous replication; switching to asynchronous replication is not possible without compromising durability, and it would not reduce CPU load on the primary. Option D is wrong: creating a read replica offloads read queries, but it does not reduce write load or CPU usage from writes on the primary instance; replication lag may persist or increase.

Answer

Increase the backup retention period to reduce I/O during backups.

Answer

Modify the DB instance to use asynchronous replication instead of synchronous.

Answer

Create a read replica in the same region and offload read queries to it.

Question 11

A company uses Amazon DynamoDB to store IoT sensor data. Each sensor writes a record every second. The table has a partition key of 'sensor_id' and a sort key of 'timestamp'. Over time, the team notices that write performance degrades for certain sensors that generate more data. The table uses provisioned capacity with auto scaling enabled. The application uses eventual consistency. The team needs to ensure consistent write performance without throttling. Which action should be taken?

Accepted Answer

Modify the partition key to include a random number or prefix to distribute writes evenly.. Adding a random number or prefix to the partition key distributes writes across multiple partitions, preventing hot partitions that cause throttling. Option B (increasing WCU) does not resolve the underlying hot partition issue, as each partition still has a maximum write capacity. Option C (on-demand mode) still has per-partition throughput limits and does not fix uneven write distribution. Option D (creating a GSI) does not affect the base table's write distribution; GSIs are for read efficiency, not write distribution.

Answer

Increase the provisioned write capacity units (WCU) to a higher value.

Answer

Switch the table to on-demand capacity mode.

Answer

Create a global secondary index (GSI) with a different partition key to handle writes.

Question 12

A company has an Amazon Redshift cluster with two dc2.large nodes. The cluster is used for daily ETL jobs and reporting. The operations team receives an alert that the cluster's disk space is 90% full. The ETL jobs are failing with 'disk full' errors. The team needs to resolve the issue quickly with minimal downtime. Which action should be taken?

Accepted Answer

Resize the cluster to a larger node type, such as dc2.large to ds2.xlarge, or add more nodes.. Resizing the cluster to a larger node type or adding more nodes increases the total storage capacity, directly addressing the disk full issue. Option A is wrong because a deep copy reorganizes data but does not increase storage capacity. Option B is wrong because VACUUM reclaims space from deleted rows but may not free enough space when the cluster is already 90% full. Option D is wrong because unloading data to S3 removes data from the cluster only if deleted, but the statement does not include deletion; even if deleted, the space is not reclaimed until VACUUM.

Answer

Perform a deep copy to re-sort and reclaim space.

Answer

Run the VACUUM command to reclaim space from deleted rows.

Answer

Unload old data to Amazon S3 and delete from the cluster.

Question 13

A company runs a production Amazon RDS for MySQL DB instance with Multi-AZ. The database is used by a web application. The application team reports that the database is experiencing intermittent connection timeouts and increased latency. The CloudWatch metrics show that the database connections spike to the maximum allowed (max_connections) during peak hours, and the CPU utilization is high. The team needs to resolve the connection issues without modifying the application code. The application uses connection pooling at the application layer. Which action should be taken?

Accepted Answer

Deploy an Amazon RDS Proxy in front of the DB instance.. RDS Proxy manages connection pooling efficiently, reducing the number of connections to the database and lowering CPU overhead. Option A is wrong because read replicas do not reduce write connection load. Option C is wrong because increasing max_connections may lead to resource exhaustion. Option D is wrong because switching to a smaller instance would worsen the problem.

Answer

Create a read replica and direct read queries to it.

Answer

Increase the 'max_connections' parameter to allow more connections.

Answer

Change the DB instance class to a smaller size to reduce the maximum connections.

Question 14

A company is migrating its on-premises Oracle database to Amazon RDS for Oracle. The database is 2 TB in size and the network bandwidth is 100 Mbps. The migration must have minimal downtime. Which TWO approaches should be used together to achieve this?

Accepted Answer

Use AWS Database Migration Service (DMS) with ongoing replication to keep the target in sync.. AWS DMS with ongoing replication allows continuous data replication with minimal downtime. AWS Snowball can be used for the initial large data transfer to avoid prolonged network transfer. Option A (taking a full backup and restoring) would cause downtime; Option B (increasing bandwidth) is not feasible quickly; Option C (SCT) is for schema conversion, not data migration; Option D (DMS with ongoing replication) is correct; Option E (Snowball for initial load then DMS for ongoing changes) minimizes network transfer time.

Answer

Take a full backup of the source database and restore it to Amazon RDS.

Answer

Increase the network bandwidth to 1 Gbps to speed up the transfer.

Answer

Use AWS Schema Conversion Tool (SCT) to convert the schema before migration.

Question 15

An e-commerce application uses Amazon Aurora MySQL with a Multi-AZ DB cluster. During a recent load test, the application experienced increased read latency. The database cluster consists of one writer and two reader instances. Which THREE actions should be taken to improve read performance?

Accepted Answer

Add more reader instances to the cluster.. Adding more reader instances (Option C) directly reduces the read load per instance, improving read performance. Enabling Aurora Auto Scaling (Option D) automatically adjusts the number of reader instances based on demand, ensuring optimal performance. Implementing Amazon ElastiCache (Option E) caches frequent queries at the application layer, reducing the number of read requests hitting the database. Option A (configuring cross-Region read replicas) introduces network latency and is primarily used for global scaling or disaster recovery, not for reducing latency in a single-region application. Option B (increasing the writer instance class) improves write performance but does not directly affect read latency.

Answer

Configure cross-Region read replicas.

Answer

Increase the instance class of the writer instance.