DBS-C01 Monitoring and Troubleshooting Practice Test 26 — 15 Questions

Question 1

A database specialist is managing an Amazon RDS for SQL Server DB instance. Users report that a critical stored procedure that previously ran in seconds now takes minutes to complete. The specialist checks Amazon CloudWatch metrics and sees that the DB instance's CPU utilization is low, but the disk queue depth is high. The specialist needs to identify the cause of the slowdown. Which action should the specialist take to diagnose the issue?

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

Accepted Answer

Query the sys.dm_os_wait_stats dynamic management view to identify the top wait types.. When CPU is low but disk queue depth is high, the bottleneck is likely disk I/O. The sys.dm_os_wait_stats DMV provides aggregated wait statistics that reveal which resources are causing delays. By identifying the top wait types, such as PAGEIOLATCH or WRITELOG, the specialist can focus on the specific queries or operations causing excessive I/O. This diagnostic step is more direct than enabling Enhanced Monitoring or resizing.

Answer

Review the SQL Server error log for disk-related errors.

Answer

Enable Enhanced Monitoring to view per-process CPU and disk I/O metrics.

Answer

Modify the DB instance class to a larger size with more IOPS.

Question 2

A database specialist is monitoring an Amazon Aurora PostgreSQL DB cluster. The cluster has a writer instance and a reader instance. The application connects to the reader endpoint for read-only queries. Users report that some read queries return stale data. The specialist needs to ensure that reads are consistent with the latest writes. Which action should the specialist take?

Accepted Answer

Configure the application to use the writer endpoint for all read queries.. Aurora PostgreSQL replicas are asynchronous by default, so reads from the reader endpoint may not include the latest committed writes. To ensure consistency, the application should direct read queries to the writer endpoint. This guarantees that all reads see the most recent data. Other options either increase lag or do not address the consistency requirement. Using the writer for reads is a simple and effective solution when consistency is critical.

Answer

Set the `hot_standby_feedback` parameter to `on` on the writer instance to prevent replication lag.

Answer

Increase the Aurora replica lag by adjusting the `max_standby_streaming_delay` parameter on the reader instance.

Answer

Enable Aurora Auto Scaling for the reader instance to reduce replica lag.

Question 3

A company uses Amazon DynamoDB with a table that has a provisioned capacity mode. The application occasionally receives ProvisionedThroughputExceededException errors, but only for certain items. A database specialist must diagnose the root cause. Which two actions should the specialist take to identify the issue? (Choose two.)

Accepted Answer

Examine the partition key design to determine if a few hot keys are causing uneven distribution of read and write activity.. ProvisionedThroughputExceededException in DynamoDB often stems from either overall capacity being too low or from hot partitions caused by an uneven partition key distribution. Checking CloudWatch consumed capacity metrics reveals if the table is hitting its provisioned limits, while analyzing the partition key design uncovers hot key patterns. Together, these actions diagnose the root cause; merely adding capacity or using unrelated services like Streams or CloudTrail will not identify why specific items are throttled.

Answer

Enable DynamoDB Streams on the table and analyze the stream records for throttled write attempts.

Answer

Use AWS CloudTrail to review the API calls that resulted in ProvisionedThroughputExceededException errors.

Answer

Increase the table's provisioned read and write capacity units immediately to resolve the exceptions.

Question 4

A database specialist is troubleshooting an Amazon Aurora PostgreSQL cluster that is experiencing replication lag on a reader instance. The application performs read-only queries on the reader endpoint. The specialist observes that the AuroraReplicaLag metric is consistently above 1000 milliseconds. The writer instance is not under heavy load. The specialist needs to reduce the replication lag. Which action should the specialist take?

Accepted Answer

Check for long-running queries on the reader that are blocking the redo apply process.. Aurora PostgreSQL replication lag on a reader can be caused by long-running queries that block the redo apply process. When a query runs for an extended period, it can prevent the apply process from advancing, leading to increased AuroraReplicaLag. The specialist should identify and terminate such queries to allow the reader to catch up. Other options like resizing or adding readers do not directly address this specific bottleneck.

Answer

Increase the instance size of the reader instance to provide more CPU and memory.

Answer

Modify the reader instance to use a larger instance class with more network bandwidth.

Answer

Enable Aurora Auto Scaling for the reader fleet to add more reader instances.

Question 5

A database specialist is monitoring an Amazon RDS for MySQL DB instance. The instance is experiencing high CPU utilization. The specialist wants to identify the top SQL statements contributing to the load using Performance Insights. (Choose two.)

Accepted Answer

Use the Performance Insights dashboard to view the Top SQL tab.. To identify top SQL statements using Performance Insights, the specialist must first enable Performance Insights on the DB instance. Then, they can use the Performance Insights dashboard and navigate to the Top SQL tab, which displays the SQL statements contributing most to database load. This combination provides the required insight into high CPU usage caused by SQL queries.

Answer

Query the MySQL performance_schema for statement digests.

Answer

Configure slow query logs and analyze them with CloudWatch Logs Insights.

Answer

Enable Enhanced Monitoring to capture OS-level metrics.

Question 6

A database specialist is monitoring an Amazon Aurora MySQL cluster. The specialist notices that the writer instance's CPU utilization is consistently high, and the application experiences increased latency. The specialist suspects that the issue is due to a high number of row lock waits. Which two actions should the specialist take to diagnose and mitigate the issue? (Choose two.)

Accepted Answer

Enable Performance Insights and analyze the db.lock_wait metric.. To diagnose row lock waits in Aurora MySQL, the specialist should enable Performance Insights to monitor the db.lock_wait metric, which quantifies the load from lock waits. Additionally, querying the performance_schema.data_lock_waits table provides detailed information about blocking and waiting transactions. These two actions together allow identification of the problematic locks and the transactions involved, enabling targeted mitigation such as query optimization or killing blocking sessions.

Answer

Increase the innodb_lock_wait_timeout parameter to reduce lock wait errors.

Answer

Query the information_schema.INNODB_LOCKS table to identify blocking transactions.

Answer

Check the MySQL error log for messages about lock wait timeouts.

Question 7

A database specialist manages an Amazon Aurora PostgreSQL cluster with a writer and a reader instance. Application logs show intermittent connection failures to the reader endpoint. The specialist suspects the reader is being restarted by Aurora. Which action will provide the most direct evidence of recent restart events on the reader instance?

Accepted Answer

Inspect the PostgreSQL log files for the reader instance via the RDS console or by downloading them from Amazon CloudWatch Logs.. Aurora PostgreSQL logs contain explicit messages for startup, shutdown, and crash recovery, providing a definitive record of restarts. CloudWatch metrics like DatabaseConnections and FreeableMemory may show symptoms but are indirect. Querying a non-existent system view is invalid. Therefore, examining the PostgreSQL logs is the most reliable way to confirm recent restart events on the reader instance.

Answer

Check the 'DatabaseConnections' metric in Amazon CloudWatch for the reader instance.

Answer

Query the 'aurora_status' view in the reader's PostgreSQL database.

Answer

Review the 'FreeableMemory' metric for the reader instance in Amazon CloudWatch.

Question 8

A database specialist is managing an Amazon RDS for SQL Server DB instance. Users report that a critical stored procedure that previously ran in seconds now takes minutes to complete. The specialist checks Amazon CloudWatch metrics and sees that the DB instance's CPU utilization is low, but the disk queue depth is high. The specialist needs to identify the cause of the slowdown. Which action should the specialist take to diagnose the issue?

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

Accepted Answer

The table has a hot partition due to an unevenly distributed partition key, causing a single partition to exceed its throughput limits.. Even in on-demand mode, DynamoDB partitions have a maximum throughput limit. A hot partition caused by an unevenly distributed partition key can lead to throttling when a single partition receives more traffic than it can handle. The solution involves designing a more uniform partition key or using techniques like write sharding to distribute the load.

Answer

The table's read capacity is set too low, and the specialist must switch to provisioned mode to set a higher capacity.

Answer

The on-demand capacity mode has a default maximum throughput of 40,000 read request units per second, and the burst exceeded that limit.

Answer

The application is using eventually consistent reads, which are throttled more aggressively than strongly consistent reads.

Question 9

A database administrator is responsible for an Amazon RDS for SQL Server DB instance. They need to monitor the performance of a specific long-running query that is causing CPU spikes. Which AWS tool provides detailed, query-level performance data with minimal impact on the DB instance?

Accepted Answer

Amazon RDS Performance Insights. Performance Insights is designed to provide query-level performance monitoring for RDS engines, including SQL Server. It captures wait events and top SQL statements by sampling active sessions, which directly identifies the query causing CPU spikes. Other tools like CloudWatch Logs, CloudTrail, or Enhanced Monitoring operate at different layers and cannot attribute CPU usage to specific queries, making Performance Insights the correct choice.

Answer

Enhanced Monitoring for RDS

Answer

Amazon CloudWatch Logs with SQL Server error logs

Answer

AWS CloudTrail with RDS API call logging

Question 10

A media company stores metadata in an Amazon DynamoDB table with on-demand capacity mode. Users report sporadic HTTP 500 errors from the application, and CloudWatch shows spikes in `ThrottledRequests` for the table. A database specialist must determine which two factors can cause on-demand tables to throttle requests even though the table is not provisioned with fixed capacity. (Choose two.)

Accepted Answer

A sudden burst of traffic exceeds double the table's previous peak throughput within a short window.. On-demand tables are not immune to throttling: they enforce per-partition throughput limits, and a hot partition from a low-cardinality key can exceed them. They also have a burst absorption limit of roughly double the previous peak, after which requests may be throttled until capacity scales. Both factors can produce sporadic `ThrottledRequests` even without provisioned capacity.

Answer

The table has DynamoDB Streams enabled, and the stream's shard iterator has expired, blocking new writes.

Answer

The table is using strongly consistent reads, which consume twice the read capacity units of eventually consistent reads.

Answer

The table's previous peak traffic was much higher than the current traffic, so DynamoDB has not yet adjusted the baseline throughput downward.

Question 11

A database specialist is investigating a production Amazon Aurora PostgreSQL cluster where the application reports intermittent connection timeouts. The specialist suspects the issue is related to the database's connection handling. The cluster uses an Aurora PostgreSQL writer instance (db.r5.xlarge) and a reader instance. The application connects to the writer endpoint. CloudWatch metrics show that DatabaseConnections is consistently near the value of max_connections, and the application logs show errors like 'FATAL: remaining connection slots are reserved for non-replication superuser connections'. What is the most likely cause of the connection timeouts?

Accepted Answer

The Aurora PostgreSQL cluster's max_connections parameter is set too low for the application's connection pool size.. The error 'remaining connection slots are reserved for non-replication superuser connections' is a definitive PostgreSQL message indicating that max_connections has been reached. CloudWatch showing DatabaseConnections near the max_connections value corroborates this. The application's connection pool is likely too large or connections are not being released properly. Increasing max_connections or implementing a connection pooler like RDS Proxy would resolve the issue.

Answer

The security group attached to the Aurora cluster is blocking the application's IP address.

Answer

The Aurora cluster's parameter group has idle_in_transaction_session_timeout set too high, causing connections to remain open.

Answer

The Aurora reader instance is not properly configured for read-only connections, causing write attempts to fail.

Question 12

A database specialist is troubleshooting an Amazon DynamoDB table that intermittently returns ThrottlingException to an application. The table uses on-demand capacity mode, and the application performs reads and writes with a mix of item sizes. The specialist wants to determine whether specific partitions are hot and whether request patterns exceed per-partition limits. (Choose two.)

Accepted Answer

Enable DynamoDB Contributor Insights for the table and review the most accessed partition keys in the report.. Contributor Insights reports the most accessed and most throttled partition keys, identifying hot keys directly. The ThrottledRequests and ReadThrottleEvents/WriteThrottleEvents metrics confirm throttling and show when it occurs. Used together, they distinguish partition-level hot spots from overall capacity pressure, which aggregate capacity metrics, streams, and auto-scaling logs cannot do for an on-demand table.

Answer

Inspect the ConsumedReadCapacityUnits and ConsumedWriteCapacityUnits metrics for the table in Amazon CloudWatch to see aggregate usage.

Answer

Turn on DynamoDB Streams with NEW_AND_OLD_IMAGES to capture the partition keys involved in throttled requests.

Answer

Use CloudWatch Logs Insights to query the table's auto-scaling logs for partition-level throttling events.

Question 13

A database specialist is troubleshooting an Amazon Aurora MySQL DB cluster that is experiencing high write latency. The specialist suspects that the issue is related to the Aurora storage subsystem. Which two actions should the specialist take to monitor and diagnose the storage-related performance? (Choose two.)

Accepted Answer

Use Amazon RDS Performance Insights to analyze DB load by wait events, specifically io/aurora_redo_log_flush.. VolumeReadTime and VolumeWriteTime provide direct measurements of Aurora storage latency, while Performance Insights wait events like io/aurora_redo_log_flush reveal storage-related waits at the database level. Together, they help diagnose whether write latency stems from the Aurora storage subsystem. Other options focus on auditing, local storage, or instance-level metrics, which are less relevant.

Answer

Monitor the FreeLocalStorage metric in CloudWatch to ensure the instance has enough local storage for writes.

Answer

Enable Enhanced Monitoring to view disk I/O metrics at the instance level.

Answer

Enable Aurora Advanced Auditing and analyze the audit logs for slow write queries.

Question 14

A company runs an Amazon ElastiCache for Redis cluster in cluster mode disabled. The operations team wants to receive an alert when the cluster's CPU utilization exceeds 80% for 5 minutes. Which AWS service should they use to create this alarm?

Accepted Answer

Amazon CloudWatch. Amazon CloudWatch is the AWS monitoring service that ingests metrics from ElastiCache, including CPUUtilization. It supports creating alarms with thresholds and evaluation periods, such as 80% for 5 minutes, and can trigger notifications via Amazon SNS. The other services serve different purposes: CloudTrail for API auditing, AWS Config for configuration compliance, and EventBridge for event routing. Thus, CloudWatch is the correct service for this alerting requirement.

Answer

Amazon EventBridge

Answer

AWS Config

Answer

AWS CloudTrail

Question 15

A database specialist is responsible for an Amazon DynamoDB table that uses on-demand capacity mode. The application experiences sudden spikes in traffic, and the specialist notices that some read requests are being throttled. The specialist wants to monitor the number of read requests that were throttled due to exceeding the table's read capacity. Which CloudWatch metric should be used?

Accepted Answer

ReadThrottleEvents. The ReadThrottleEvents metric specifically counts read requests that DynamoDB throttled due to exceeding the table's read capacity. In on-demand mode, throttling can occur if the request rate exceeds the previous peak by more than double. Monitoring ReadThrottleEvents allows the specialist to detect and respond to throttling, such as by adjusting capacity mode or optimizing queries. This metric is the direct indicator for read throttling.

Answer

ThrottledRequests

Answer

ConsumedReadCapacityUnits

Answer

SuccessfulRequestLatency