DBS-C01 Monitoring and Troubleshooting Practice Test 15 — 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 has a local secondary index (LSI) on 'OrderStatus' and a global secondary index (GSI) on 'ProductCategory'. During a flash sale, the application experiences throttling on the GSI, but the base table's provisioned throughput is not fully utilized. CloudWatch metrics show high ConsumedWriteCapacityUnits on the GSI but low on the base table. What is the most likely cause of the throttling?

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

Accepted Answer

The GSI's partition key 'ProductCategory' has a low cardinality, causing hot partitions.. The GSI's partition key 'ProductCategory' likely has few distinct values, leading to hot partitions. Write operations for many items with the same ProductCategory value are directed to a single partition, exceeding its throughput limit. This causes throttling on the GSI even though the base table has spare capacity because the base table's partition key 'CustomerID' distributes writes more evenly.

Answer

The GSI's write capacity is not sufficient because it is provisioned independently from the base table.

Answer

The GSI's sort key is missing, causing all writes to go to a single partition.

Answer

The LSI on 'OrderStatus' is consuming write capacity from the base table, causing throttling on the GSI.

Question 2

A database specialist is managing an Amazon DynamoDB table that uses on-demand capacity mode. The application performs frequent conditional writes, and the specialist notices that the table's consumed write capacity is much higher than expected based on the number of items written. The specialist needs to identify the cause of the increased capacity consumption. Which factor is most likely responsible?

Accepted Answer

The conditional writes are failing and being retried, leading to multiple write attempts for each logical write.. In DynamoDB, conditional writes that fail still consume write capacity. When an application uses conditional writes and retries on failure, each retry consumes additional WCUs. This can lead to consumed write capacity being much higher than the number of successful writes. The specialist should review the application's retry logic and condition expressions to minimize unnecessary consumption.

Answer

The table's items are larger than 1 KB, causing each write to consume multiple write capacity units.

Answer

The table is configured with a local secondary index (LSI) that shares the same partition key, causing write amplification.

Answer

The table has a global secondary index (GSI) that is not projected with all attributes, causing additional reads during writes.

Question 3

A database specialist is troubleshooting an Amazon Aurora MySQL cluster that experiences occasional performance degradation. The cluster has a writer instance and a reader instance. The specialist notices that the writer instance's CPU utilization spikes to 100% for short periods, and during these spikes, the reader instance's replica lag increases significantly. The application uses the reader endpoint for read-only queries. What is the most likely cause of the increased replica lag during writer CPU spikes?

Accepted Answer

The writer instance is unable to ship binary logs to the reader instance due to CPU contention, causing replication delay.. In Amazon Aurora MySQL, the writer instance is responsible for generating and shipping redo log records to reader instances. When the writer's CPU is saturated, it may not have enough CPU cycles to promptly ship these logs, leading to increased replica lag on the readers. This is a common cause of transient replica lag during write-heavy periods or CPU-intensive operations on the writer. Monitoring both writer CPU and replica lag is essential to diagnose such issues.

Answer

The reader instance is using a smaller instance class than the writer, making it unable to keep up with the writer's transaction volume.

Answer

The reader instance is overwhelmed by read queries from the application, causing it to fall behind on applying replication logs.

Answer

The Aurora cluster's replication is synchronous, so the writer waits for the reader to acknowledge each write, causing lag during CPU spikes.

Question 4

A database specialist is monitoring an Amazon RDS for SQL Server DB instance. Users report that a specific query has become very slow. The specialist suspects blocking. Which two actions should the specialist take to identify the blocking session and the blocking query? (Choose two.)

Accepted Answer

Use SQL Server Profiler to trace the Blocked Process Report event.. To identify blocking, querying sys.dm_exec_requests for blocking_session_id reveals the blocking session, and then you can get the query from sys.dm_exec_sql_text. Alternatively, SQL Server Profiler with the Blocked Process Report event captures detailed blocking reports including T-SQL statements. These two methods directly provide both the blocking session and the query. Enhanced Monitoring and storage metrics lack session-level detail, and sp_who2 does not provide the blocking query text.

Answer

Enable and query the RDS Enhanced Monitoring metrics for CPU utilization.

Answer

Run sp_who2 to list all active sessions and look for BlkBy values.

Answer

Check the RDS console for the DB instance's ReadIOPS and WriteIOPS metrics.

Question 5

A company uses Amazon DynamoDB for a critical application. A database specialist notices that during peak hours, the application experiences increased latency and occasional throttling. The DynamoDB table is configured with provisioned capacity mode, and the specialist reviews the CloudWatch metrics: ConsumedReadCapacityUnits and ConsumedWriteCapacityUnits are consistently near the provisioned limits, but not exceeding them. However, ThrottledRequests metric shows occasional spikes. The table has a global secondary index (GSI) with its own provisioned capacity. What is the most likely cause of the throttling?

Accepted Answer

The GSI's provisioned capacity is insufficient for the write workload, causing throttling on the base table.. When a DynamoDB table has a global secondary index, writes to the base table also consume write capacity on the GSI. If the GSI's provisioned write capacity is insufficient, the write operation can be throttled even if the base table has available capacity. The CloudWatch metrics for the base table may show consumed capacity near the provisioned limit, but the throttling is actually caused by the GSI's capacity. To resolve, increase the GSI's provisioned write capacity or switch to on-demand mode.

Answer

The DynamoDB table is experiencing hot partitions, causing uneven capacity consumption and throttling.

Answer

The application is using eventually consistent reads, which consume more capacity than strongly consistent reads, leading to throttling.

Answer

The base table's provisioned capacity is too low, causing throttling despite the metrics showing consumed capacity near the limit.

Question 6

A company uses Amazon ElastiCache for Redis to cache session data. The operations team reports that the cache hit rate has dropped from 95% to 60% over the past week. The Redis cluster shows no errors, and memory usage is at 70%. The team needs to determine why the hit rate has decreased. Which action should they take first?

Accepted Answer

Review the Redis keyspace and check for an increase in evictions or expired keys.. A significant drop in cache hit rate without errors or memory pressure often points to keys being evicted or expiring. In ElastiCache for Redis, the evicted_keys and expired_keys metrics in CloudWatch can indicate if keys are being removed. With memory at 70%, eviction is unlikely, so expired keys due to TTL settings are a probable cause. Reviewing the keyspace and TTL configurations is the correct first step.

Answer

Enable cluster mode and reshard the data to improve cache locality.

Answer

Increase the number of shards in the Redis cluster to distribute the load.

Answer

Check the application logs to see if the application is requesting different keys than before.

Question 7

A database specialist is monitoring an Amazon Aurora PostgreSQL DB cluster. The cluster consists of one writer and one reader. The specialist observes that the reader instance is experiencing high replica lag, and the application reading from the reader is seeing stale data. The specialist needs to identify the causes of the replica lag. (Choose two.)

Accepted Answer

The writer instance is experiencing high write throughput, causing the reader to fall behind in applying changes.. Replica lag in Aurora PostgreSQL can be caused by high write throughput on the writer that the reader cannot keep up with, or by the reader being under-provisioned relative to the workload. Both factors directly affect the reader's ability to apply changes from the shared storage volume. Other options are invalid because they reference non-existent features or misunderstandings of Aurora's architecture.

Answer

The writer instance is performing a large bulk load operation, generating a high volume of redo log records.

Answer

The reader instance is using a different storage type than the writer, causing I/O bottlenecks.

Answer

The Aurora cluster's parameter group has hot standby enabled, which delays replication to the reader.

Question 8

A database specialist is troubleshooting an Amazon Aurora PostgreSQL cluster that is experiencing performance degradation. The specialist observes that the 'BufferCacheHitRatio' metric in Amazon CloudWatch is low, and the 'FreeableMemory' metric is fluctuating. The cluster has one writer and two readers. The application uses the reader endpoint for read-only queries. Which two actions should the specialist take to improve performance? (Choose two.)

Accepted Answer

Enable the Aurora Auto Scaling feature to automatically add reader instances based on workload.. A low BufferCacheHitRatio on Aurora PostgreSQL indicates that the buffer cache is not effectively caching data, leading to increased disk I/O and slower performance. To address this, increasing the memory available to the reader instances by scaling them up allows more data to be cached, improving the hit ratio. Additionally, enabling Aurora Auto Scaling can dynamically add reader instances to distribute the read load, reducing the pressure on each instance and further improving cache efficiency. Both actions target the read workload, which is handled by the reader instances.

Answer

Modify the Aurora cluster to use a larger instance class for the writer instance.

Answer

Change the Aurora cluster to use a larger instance class for the reader instances and enable the query cache.

Answer

Increase the 'max_connections' parameter for the Aurora cluster to allow more concurrent connections.

Question 9

A company uses Amazon DocumentDB (with MongoDB compatibility) and has enabled change streams to capture data changes. The application consumes change stream events from a tailable cursor. Recently, the application has been experiencing increased latency in processing events, and the database specialist notices that the cursor is frequently returning empty batches. The specialist needs to ensure that the application can keep up with the change stream and reduce latency. Which action should the specialist take?

Accepted Answer

Increase the number of application consumers reading from the change stream and distribute the load.. The empty batches suggest the consumer is polling faster than events are being produced or processed. To reduce latency and keep up, scaling out consumers is effective. Each consumer can have its own change stream cursor, and together they process events in parallel. Increasing batch size or scaling up the database does not address consumer-side bottlenecks. The resume token cache is not configurable. Thus, adding consumers is the appropriate action.

Answer

Scale up the DocumentDB instance class to provide more CPU and memory for change stream processing.

Answer

Configure the change stream to use a larger resume token cache.

Answer

Increase the change stream cursor's batch size by setting the batchSize option to a higher value.

Question 10

A database administrator is monitoring an Amazon DynamoDB table using Amazon CloudWatch. The administrator wants to receive an alert when the table's consumed write capacity exceeds 80% of the provisioned capacity for 5 consecutive minutes. Which CloudWatch metric and configuration should the administrator use?

Accepted Answer

Create a CloudWatch alarm using a math expression that calculates the percentage of consumed write capacity relative to provisioned capacity, with a threshold of 80% and a period of 5 minutes.. To alert when consumed write capacity exceeds 80% of provisioned capacity for 5 consecutive minutes, the administrator should use a CloudWatch math expression. This expression can calculate the percentage by dividing ConsumedWriteCapacityUnits by the provisioned capacity and multiplying by 100. Setting the alarm threshold to 80 with a period of 5 minutes and evaluation periods of 1 will trigger when the condition is met. This method is accurate and automatically adjusts if the provisioned capacity changes.

Answer

Create a CloudWatch alarm on the ConsumedWriteCapacityUnits metric with a threshold of 80% of the provisioned capacity and a period of 5 minutes.

Answer

Create a CloudWatch alarm on the ConsumedWriteCapacityUnits metric with a threshold equal to 80% of the provisioned capacity and a period of 1 minute, then use a composite alarm to require 5 consecutive data points.

Answer

Create a CloudWatch alarm on the WriteThrottleEvents metric with a threshold greater than 0 for 5 consecutive minutes.

Question 11

A company is using Amazon RDS for MySQL and notices that the Read IOPS metric is consistently high during business hours. The application is read-heavy. Which configuration change would most likely reduce Read IOPS?

Accepted Answer

Create one or more read replicas and redirect read traffic to them.. Creating read replicas offloads read queries from the primary DB instance to replica instances, directly reducing the number of read I/O operations on the primary. Since the application is read-heavy and Read IOPS is high during business hours, distributing read traffic to replicas alleviates the I/O bottleneck on the primary instance without requiring a larger instance or storage changes.

Answer

Add a Multi-AZ standby instance.

Answer

Increase the DB instance size to a larger instance type.

Answer

Enable storage Auto Scaling on the RDS instance.

Question 12

A developer reports that an application using Amazon DynamoDB is experiencing high latency during peak hours. The table has a provisioned capacity of 500 read capacity units (RCUs) and 500 write capacity units (WCUs). The application uses eventually consistent reads and the table is about 50 GB. The developer notices throttled write requests in CloudWatch. Which action would most effectively reduce write throttling?

Accepted Answer

Increase the provisioned write capacity for the table.. The developer reports throttled write requests, which directly indicates that the provisioned write capacity (500 WCUs) is insufficient to handle the peak write traffic. Increasing the provisioned write capacity for the table is the most direct and effective action to eliminate write throttling, as it raises the limit on write operations per second. Option C is correct because it addresses the root cause—write capacity exhaustion—without introducing unnecessary components or changing read behavior.

Answer

Enable DynamoDB Accelerator (DAX) for the table.

Answer

Create a global secondary index on the table.

Answer

Switch from eventually consistent reads to strongly consistent reads.

Question 13

A company is migrating an on-premises Oracle database to Amazon RDS for Oracle. During the migration, the database administrator notices that the CPU utilization on the RDS instance is consistently above 90% during peak hours, even though the on-premises server had similar specifications. The application queries are mostly SELECT statements with occasional DML. The RDS instance is db.r5.large with 500 GB of General Purpose SSD (gp2) storage. Which change would most likely reduce CPU utilization?

Accepted Answer

Upgrade to a larger instance type, such as db.r5.xlarge.. The db.r5.large instance type has 2 vCPUs and 16 GiB of memory. Sustained CPU utilization above 90% during peak hours indicates that the instance is compute-bound for the workload. Upgrading to db.r5.xlarge (4 vCPUs, 32 GiB memory) doubles the available CPU capacity, directly reducing CPU utilization for the same query load. The on-premises server had similar specifications, but RDS instances may have different CPU architectures or hypervisor overhead, making the larger instance the most direct fix.

Answer

Create a read replica and redirect all SELECT queries to the replica.

Answer

Enable Multi-AZ to offload CPU to the standby instance.

Answer

Increase the allocated storage to 1 TB to improve I/O performance.

Question 14

A team manages an Amazon Aurora MySQL database. They observe that the 'Deadlocks' metric in CloudWatch is spiking. The application uses a single writer instance and multiple read replicas. Which action is most effective at reducing deadlocks?

Accepted Answer

Review application code to ensure transactions are as short as possible and access tables in a consistent order.. Deadlocks in Aurora MySQL occur when two or more transactions hold locks that the other needs, and they wait indefinitely. The most effective way to reduce deadlocks is to keep transactions short and access tables in a consistent order, which minimizes lock contention and avoids circular wait conditions. This directly addresses the root cause of deadlocks, unlike scaling or redirecting traffic, which only reduce the probability of contention without fixing the underlying locking pattern.

Answer

Increase the instance size to handle more concurrent connections.

Answer

Redirect read traffic to read replicas to reduce load on the writer.

Answer

Enable Multi-AZ to distribute the load.

Question 15

A database engineer is troubleshooting slow query performance on an Amazon RDS for PostgreSQL instance. The instance is db.r5.large with 500 GB of General Purpose SSD (gp2) storage. CloudWatch metrics show high Read Latency and high Read IOPS, but low CPU utilization. Which TWO actions should the engineer take to improve performance?

Accepted Answer

Create a read replica and offload read queries to it.. A is correct because creating a read replica offloads read queries from the primary instance, reducing the read IOPS and read latency on the primary. This directly addresses the high Read Latency and high Read IOPS metrics without requiring a larger instance class or storage change, especially since CPU utilization is low, indicating the bottleneck is I/O, not compute.

Answer

Increase the DB instance class to a larger size, such as db.r5.2xlarge.

Answer

Enable Multi-AZ to use the standby for read traffic.

Answer

Optimize queries by adding appropriate indexes.