DBS-C01 Monitoring and Troubleshooting Practice Test 25 — 15 Questions

Question 1

A company uses Amazon DynamoDB with a table that has a partition key of 'CustomerID' and a sort key of 'OrderDate'. The table is configured with on-demand capacity mode. During a marketing campaign, the application experiences a sudden increase in write traffic. The operations team notices that some write requests are being throttled with 'ProvisionedThroughputExceededException' errors, even though the table is in on-demand mode. Which of the following is the MOST likely explanation for these throttling errors?

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

Accepted Answer

The write traffic is concentrated on a few partition key values, causing a hot partition that exceeds the per-partition throughput limit.. In DynamoDB on-demand mode, throttling can still occur if a single partition exceeds its throughput limit. This happens when write traffic is concentrated on a few partition key values, creating a hot partition. The other options are less likely given the scenario.

Answer

The table's sort key 'OrderDate' is causing an uneven distribution of data across partitions, leading to throttling.

Answer

The on-demand capacity mode requires a warm-up period after a sudden traffic increase, during which throttling can occur.

Answer

The table's on-demand capacity mode has a default maximum throughput limit that is being exceeded.

Question 2

A company uses Amazon ElastiCache for Redis to cache session data for a web application. The Redis cluster is in cluster mode disabled with one shard and two replicas. Users are reporting that they are being logged out unexpectedly. A database specialist checks the ElastiCache metrics and notices that the CurrConnections metric is very high and the Evictions metric is increasing. The cache node type is cache.m5.large with 6.38 GiB of memory. What is the MOST likely cause of the session loss?

Accepted Answer

The Redis cluster's maxmemory-policy is set to allkeys-lru, and the cache is evicting session keys due to memory pressure.. The combination of high CurrConnections and increasing Evictions indicates that the Redis instance is running out of memory and evicting keys to make room. With a policy like allkeys-lru, session keys can be evicted, causing users to be logged out. To resolve this, consider increasing the node size, using a more appropriate eviction policy such as volatile-lru if sessions have TTLs, or optimizing memory usage.

Answer

The application is not using Redis correctly; it should store session data in a Redis sorted set instead of strings.

Answer

The Redis cluster is experiencing a failover, causing the primary node to change and session data to be lost.

Answer

The ElastiCache cluster's parameter group has maxmemory-policy set to noeviction, causing writes to fail and sessions to be lost.

Question 3

A database specialist is troubleshooting an Amazon Aurora PostgreSQL cluster where the application reports intermittent connection timeouts during peak hours. The specialist suspects a connection storm and wants to identify the source. Which metric should be monitored in Amazon CloudWatch to determine the number of current connections to the DB instance?

Accepted Answer

DatabaseConnections. The DatabaseConnections metric in Amazon CloudWatch reports the number of client connections to the DB instance. In an Aurora PostgreSQL cluster, monitoring this metric is essential to detect connection storms that can lead to timeouts. By tracking DatabaseConnections, the specialist can determine if the application is exceeding connection limits or if there is a sudden surge in connections, enabling appropriate mitigation such as using connection pooling or increasing max_connections.

Answer

DBLoad

Answer

LoginFailures

Answer

NetworkReceiveThroughput

Question 4

A database specialist manages an Amazon DynamoDB table that uses on-demand capacity mode. The application experiences sudden spikes in write traffic, and the specialist observes that some write requests are being throttled with `ThrottlingException` errors. The specialist needs to identify the cause of the throttling and ensure that the application can handle the spikes without errors. Which action should the specialist take?

Accepted Answer

Request a service quota increase for the table's on-demand write throughput limit.. In on-demand mode, DynamoDB automatically scales but is constrained by a default per-table throughput quota. Sudden spikes exceeding this quota cause throttling. Requesting a quota increase raises the limit, allowing the table to handle higher throughput without errors, which is the correct immediate action.

Answer

Enable DynamoDB Accelerator (DAX) for the table to offload write traffic.

Answer

Modify the table to use provisioned capacity mode with auto scaling enabled.

Answer

Increase the table's provisioned write capacity units (WCUs) to accommodate the peak traffic.

Question 5

A database specialist is troubleshooting an Amazon Aurora PostgreSQL DB cluster that is experiencing performance degradation. The cluster has one writer and one reader instance. The specialist observes that the writer's CPU utilization is consistently above 90%, while the reader's CPU is around 30%. The application performs a mix of read and write operations. The specialist wants to identify the cause of the high CPU on the writer. Which TWO actions should the specialist take to diagnose the issue? (Choose two.)

Accepted Answer

Review the PostgreSQL pg_stat_statements view on the writer instance to identify queries with high total execution time.. To diagnose high CPU on the Aurora PostgreSQL writer, the specialist should use tools that identify resource-intensive queries. Performance Insights provides a visual dashboard of database load by SQL statement, and pg_stat_statements offers detailed query execution statistics. Both directly reveal which queries are consuming the most CPU, enabling targeted optimization. Other options focus on replication, disk I/O, or connection errors, which are less relevant to the observed CPU issue.

Answer

Examine the Enhanced Monitoring metrics for the writer instance to check for high disk I/O wait.

Answer

Check the Aurora Replica lag metric to determine if the reader is falling behind.

Answer

Use Amazon CloudWatch Logs Insights to analyze the PostgreSQL error logs for frequent connection errors.

Question 6

A database specialist manages an Amazon Aurora PostgreSQL DB cluster. Users report that queries against a reader instance intermittently return stale data that is several seconds behind the writer. The specialist wants to quantify this lag over time and identify whether it correlates with writer activity. Which AWS mechanism should the specialist use to obtain this information?

Accepted Answer

Enable Aurora Replica lag monitoring by publishing the AmazonAuroraReplicaLag metric from the DB cluster to Amazon CloudWatch.. Aurora publishes AmazonAuroraReplicaLag in milliseconds for each Aurora Replica to CloudWatch. Graphing that metric over time is the supported way to quantify how far behind a reader is and to correlate lag with writer activity. The other mechanisms either report unrelated performance data or rely on PostgreSQL streaming-replication views that do not reflect Aurora storage-level replication.

Answer

Query the pg_stat_replication view on the writer instance to read the replay_lag column for each connected replica.

Answer

Enable Performance Insights on the reader instance and inspect the db.sql.statement metric for delayed queries.

Answer

Enable Enhanced Monitoring on the DB cluster and read the replica_lag metric from the OS-level process list.

Question 7

A database specialist is monitoring an Amazon RDS for SQL Server DB instance. The instance is experiencing intermittent connection timeouts. The specialist suspects that the issue is related to the maximum number of worker threads or the number of user connections. Which two metrics should the specialist review in Amazon CloudWatch to confirm the cause? (Choose two.)

Accepted Answer

DatabaseConnections. Connection timeouts in RDS for SQL Server can result from exhausting the maximum user connections or the maximum worker threads. DatabaseConnections shows current client connections, and Threads shows active threads. Together they indicate whether either limit is being approached. CPUUtilization and FreeableMemory are general health metrics, and UserConnections is not a valid RDS metric.

Answer

CPUUtilization

Answer

UserConnections

Answer

FreeableMemory

Question 8

A database specialist manages an Amazon Aurora MySQL cluster used for an online ordering system. During peak traffic, the application reports occasional connection timeouts and slow queries. The specialist enables Performance Insights with a retention period of 7 days and observes that the wait event 'io/table/sql/handler' dominates during those periods. The cluster has one writer and one reader, both using db.r5.2xlarge instances. Which action should the specialist take to reduce the impact of this wait event?

Accepted Answer

Analyze the top SQL statements in Performance Insights and add appropriate indexes to reduce the number of rows scanned.. The dominant wait event 'io/table/sql/handler' points to I/O bottlenecks caused by inefficient SQL. Performance Insights provides detailed metrics to identify the top SQL statements contributing to the load. By analyzing these statements and adding suitable indexes or rewriting queries to reduce logical and physical reads, the specialist can lower I/O waits and improve response times. Instance scaling or parallel query may help in specific cases but do not directly resolve the underlying query inefficiency.

Answer

Enable the Aurora parallel query feature on the cluster to offload I/O processing to the storage layer.

Answer

Increase the size of the Aurora cluster's buffer pool by scaling the writer instance to db.r5.4xlarge.

Answer

Modify the Aurora cluster to use a larger instance class for the reader and redirect all read traffic to the reader.

Question 9

A database specialist is monitoring an Amazon DocumentDB cluster. The application reports slow query performance. The specialist checks the cluster's metrics and sees that 'BufferCacheHitRatio' is below 90% and 'DiskQueueDepth' is high. Which action should the specialist take to improve performance?

Accepted Answer

Increase the instance class to a larger size to provide more memory for the buffer cache.. BufferCacheHitRatio below 90% and high DiskQueueDepth indicate that the instance lacks sufficient memory to cache frequently accessed data. Increasing the instance class provides more RAM, enlarging the buffer cache and reducing disk reads. This directly improves query performance by allowing more data to be served from memory.

Answer

Increase the cluster's storage volume size to improve IOPS.

Answer

Enable Performance Insights to identify the slow queries.

Answer

Add a read replica to distribute the read workload.

Question 10

A database administrator is responsible for an Amazon RDS for MySQL DB instance. The administrator needs to be alerted when the free storage space falls below 10% of the allocated storage. The administrator wants to receive an email notification. What is the MOST efficient way to achieve this?

Accepted Answer

Create a CloudWatch alarm on the FreeStorageSpace metric with a threshold of 10% of allocated storage, and configure an SNS topic to send email notifications.. CloudWatch alarms are the standard way to monitor RDS metrics and trigger actions. The FreeStorageSpace metric is available out-of-the-box. By setting an alarm threshold at 10% of allocated storage and linking it to an SNS topic with an email subscription, the administrator gets notified automatically. This requires minimal configuration and no custom code.

Answer

Use AWS Trusted Advisor to monitor RDS storage and configure an email notification.

Answer

Enable RDS Enhanced Monitoring and create a CloudWatch Logs subscription filter to send alerts.

Answer

Create an AWS Lambda function that queries the RDS instance for free storage space and sends an email via Amazon SES.

Question 11

A database specialist is troubleshooting an Amazon RDS for SQL Server DB instance that is experiencing performance degradation. The specialist runs a query against the `sys.dm_exec_requests` dynamic management view and notices a high number of sessions with a wait type of `PAGEIOLATCH_SH`. The specialist needs to identify the cause and resolve the issue. Which action should the specialist take?

Accepted Answer

Review the instance's memory configuration and consider increasing the DB instance class to add more RAM.. PAGEIOLATCH_SH waits indicate SQL Server is waiting for data pages to be read from disk into memory. The most common cause is insufficient memory, causing frequent buffer pool evictions. Increasing the DB instance class to add RAM expands the buffer pool, reducing physical I/O and alleviating these waits.

Answer

Enable and analyze the slow query log to identify long-running queries causing page reads.

Answer

Increase the allocated storage for the DB instance to reduce I/O latency.

Answer

Modify the DB parameter group to increase the `max degree of parallelism` (MAXDOP) setting.

Question 12

A company uses Amazon DynamoDB with a table that has a partition key of 'CustomerID' and a sort key of 'OrderDate'. The table is configured with on-demand capacity mode. During a marketing campaign, the application experiences a sudden increase in write traffic. The operations team notices that some write requests are being throttled with 'ProvisionedThroughputExceededException' errors, even though the table is in on-demand mode. Which of the following is the MOST likely explanation for these throttling errors?

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

Accepted Answer

The GSI's write capacity is too low because every update to the GSI key attribute results in a delete and an insert in the GSI, consuming twice the write capacity units.. Updating a GSI key attribute requires DynamoDB to delete the existing index entry and insert a new one. Each operation consumes write capacity units from the GSI. This effectively doubles the write cost for that update. Since the GSI has only 100 WCU, it can become throttled even if the base table has sufficient capacity. Increasing the GSI's write capacity resolves the bottleneck.

Answer

The GSI is throttled because DynamoDB replicates the entire item to the GSI on every update, consuming write capacity proportional to the item size, not just the key attributes.

Answer

The GSI is throttled because the base table's write capacity is consumed by the GSI updates, leaving insufficient capacity for the base table, and the throttling is misattributed to the GSI.

Answer

The GSI's write capacity is not the issue; DynamoDB automatically scales GSI capacity based on the base table's capacity, so the throttling must be due to a hot partition in the base table.

Question 13

A company runs an Amazon RDS for MySQL Multi-AZ DB instance. During a maintenance window, the DB instance fails over to the standby. After failover, the application's connections through the original DNS endpoint fail for about a minute, then recover automatically. A database specialist must explain why the application recovered without any endpoint change. What is the reason?

Accepted Answer

The CNAME record of the DB instance endpoint is updated to point to the new primary, and clients that honor the record's short TTL reconnect to the new primary.. In a Multi-AZ DB instance deployment, failover promotes the standby and updates the DB instance endpoint's DNS CNAME to the new primary. Applications that reconnect by resolving the endpoint name recover automatically once the record's TTL expires and DNS propagates. This is why no endpoint configuration change is required on the client side, and why a brief connection interruption is expected.

Answer

The application uses the cluster endpoint, which automatically redirects connections to whichever node is currently the writer.

Answer

RDS Proxy maintains persistent client connections and transparently reconnects them to the new primary after failover.

Answer

The standby assumes the original primary's private IP address during failover, so existing connections continue to work without re-resolution.

Question 14

A database specialist is troubleshooting an Amazon ElastiCache for Redis cluster that is experiencing increased latency. The cluster uses cluster mode disabled and has one shard with two replicas. The specialist notices that the engine CPU utilization is low, but the database is reporting many new connections per second. Which action should the specialist take to reduce latency?

Accepted Answer

Configure the application to use a connection pool or persistent connections to reuse existing connections.. A high number of new connections per second indicates connection churn, where clients repeatedly open and close connections. This overhead consumes server resources and increases latency. The most effective solution is to implement connection pooling or persistent connections in the application. This reduces the rate of new connections and improves overall performance. Scaling up or adding replicas does not fix the client behavior.

Answer

Scale up to a larger node type to handle the connection rate.

Answer

Increase the number of replicas to distribute the connection load.

Answer

Enable encryption in transit to reduce connection overhead.

Question 15

A company uses Amazon DynamoDB with a table that has a partition key of 'CustomerID' and a sort key of 'OrderDate'. The table is configured with on-demand capacity mode. During a marketing campaign, the application experiences a sudden increase in write traffic. The operations team notices that some write requests are being throttled with 'ProvisionedThroughputExceededException' errors, even though the table is in on-demand mode. Which of the following is the MOST likely explanation for these throttling errors?

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

Accepted Answer

Amazon CloudWatch Contributor Insights for DynamoDB. CloudWatch Contributor Insights for DynamoDB is specifically designed to identify the most accessed and throttled partition keys by analyzing DynamoDB logs. It provides ranked lists and graphs in near real-time, allowing the team to quickly find hot keys during the flash sale. Other options either do not capture read capacity consumption per key or require significant custom effort without the same level of insight.

Answer

DynamoDB Accelerator (DAX) metrics in CloudWatch

Answer

Amazon DynamoDB Streams with an AWS Lambda function to log item-level changes

Answer

AWS CloudTrail data events for the DynamoDB table