Courseiva

CCNA Troubleshooting Questions

75 of 289 questions · Page 3/4 · Troubleshooting topic · Answers revealed

151
MCQmedium

A company uses Amazon DynamoDB to store sensor data from IoT devices. The table has a partition key of DeviceID (String) and a sort key of Timestamp (Number). The table is configured with provisioned capacity of 1000 read capacity units (RCUs) and 500 write capacity units (WCUs). Auto Scaling is enabled with target utilization of 70% and is working correctly. Recently, a new batch job was introduced that performs Scan operations on the entire table every hour. The Scan consumes many read capacity units and causes throttling of the sensor write requests. The team wants to minimize the impact on the write-heavy sensor ingestion. Which approach is BEST?

A.Increase the provisioned write capacity to 2000 WCUs to handle bursts.
B.Schedule the Scan to run during off-peak hours when sensor writes are lower.
C.Enable DAX (DynamoDB Accelerator) to cache Scan results and reduce read consumption on the table.
D.Switch the table to on-demand capacity mode to eliminate throttling.
AnswerD

On-demand capacity mode automatically scales to handle any traffic burst, eliminating throttling for both reads and writes. It is the best choice to prevent scan-induced throttling of write requests.

Why this answer

Switching to on-demand capacity mode eliminates throttling by automatically scaling read/write capacity to meet traffic demands. Since the scan operation runs hourly and consumes significant RCUs, on-demand mode will handle the burst without throttling write requests. Option C is incorrect because DAX caches individual items, not entire scan results; hourly scans would cause cache misses, providing no benefit.

Option A is incorrect because increasing write capacity does not address read throttling. Option B is incorrect because off-peak scheduling may not be feasible and does not guarantee elimination of throttling.

152
MCQmedium

A company is running a MongoDB-compatible Amazon DocumentDB cluster. The application team reports that write operations are failing intermittently with a `WriteConcernError` indicating that the write concern could not be satisfied. The cluster has one primary and two replicas. What is the MOST likely cause of this issue?

A.One of the replicas is down or experiencing high replication lag
B.The cluster does not have enough replica instances to satisfy the write concern
C.The application is using an incorrect read preference
D.The primary instance is overloaded with read requests
AnswerA

If a replica is down, the write concern 'majority' cannot be satisfied because only the primary and one replica are available, but majority may require two replicas depending on configuration.

Why this answer

The WriteConcernError occurs because the write concern requires acknowledgment from a certain number of replicas. With a cluster of one primary and two replicas, if any replica is down or has high replication lag, the write concern requirement cannot be satisfied. Option A is correct because that is the most likely cause.

Option B is incorrect because the cluster has enough replicas for majority write concern if all are healthy. Option C is incorrect because read preference does not affect write operations. Option D is incorrect because the primary being overloaded with reads would not cause a WriteConcernError; it would affect performance but not the ability to satisfy write concern.

153
Multi-Selecthard

A company is migrating an on-premises SQL Server database to Amazon RDS for SQL Server. During the migration, the DBA notices that the network latency between the source and target is high. Which THREE actions can the DBA take to reduce the impact of network latency on migration performance?

Select 3 answers
A.Increase the size of the RDS instance
B.Use multiple parallel tasks in AWS DMS
C.Use AWS Direct Connect for a dedicated network connection
D.Enable data compression for the migration task
E.Use AWS Site-to-Site VPN
AnswersB, C, D

Parallel tasks improve throughput, mitigating latency effects.

Why this answer

Options B, C, and D are correct. Using multiple parallel tasks in AWS DMS (option B) improves throughput by transferring data in parallel. Using AWS Direct Connect (option C) provides a dedicated, low-latency network connection.

Enabling data compression (option D) reduces the amount of data transferred, mitigating latency impact. Option A is incorrect because increasing the RDS instance size does not reduce network latency; it only affects database processing capacity. Option E is incorrect because using a VPN can add overhead and may not reduce latency.

154
MCQhard

A company is using Amazon DynamoDB with global tables to support a multi-region application. They notice that writes to the table in us-east-1 are not replicating to the table in eu-west-1. The replication lag metric shows increasing values. What is the most likely cause?

A.The table in eu-west-1 has a lower write capacity and is throttling incoming writes.
B.DynamoDB Streams is disabled on the table in us-east-1.
C.The DynamoDB table in us-east-1 is not accessible from eu-west-1 due to VPC restrictions.
D.The table in eu-west-1 has been removed as a replica region from the global table configuration.
AnswerD

If the replica region is removed, replication stops.

Why this answer

If the replica region (eu-west-1) has been removed from the global table configuration, DynamoDB will stop replicating writes to that region, and the replication lag metric will show increasing values. Option A is incorrect: while a lower write capacity in eu-west-1 could cause throttling and replication lag, it would not completely stop replication; the lag metric would still show some activity. Option B is incorrect: DynamoDB Streams must be enabled on the source table for global tables to function, but disabling them would halt replication entirely, not cause increasing lag.

Option C is incorrect: global tables replicate asynchronously over DynamoDB's internal network and do not require VPC connectivity between regions.

155
MCQhard

A company runs a critical application on Amazon RDS for PostgreSQL with Multi-AZ deployment. The database is a db.r5.xlarge instance with 500 GB of Provisioned IOPS (io1) storage. The application writes a large volume of time-series data every 5 minutes. Recently, the team noticed that the replica lag between the primary and standby instance has increased to over 10 seconds during peak write periods. The application requires that the standby be within 5 seconds of the primary for disaster recovery purposes. The team has already verified that network latency between AZs is minimal, and the standby instance is the same size as the primary. The CloudWatch metric 'TransactionLogsDiskUsage' shows that the primary is generating logs at a high rate. Which combination of actions should the database administrator take to reduce the replica lag? (Choose the best option.)

A.Disable synchronous commit on the primary to reduce log generation
B.Change the standby instance to a larger instance type (e.g., db.r5.2xlarge) to improve apply throughput
C.Reduce the batch write frequency to every 10 minutes to lower the log generation rate
D.Increase the Provisioned IOPS on both the primary and standby instances to improve log write throughput
AnswerD

Higher IOPS reduces the time to flush transaction logs, reducing replica lag.

Why this answer

Increasing Provisioned IOPS on both the primary and standby instances improves I/O throughput for writing transaction logs, which is the bottleneck causing replica lag. Higher IOPS reduces the time needed to generate logs on the primary and apply them on the standby. Option A is incorrect because disabling synchronous commit would reduce durability and may still not achieve the required lag reduction.

Option B is incorrect since the standby instance size is already equal to the primary; the issue is I/O performance, not compute capacity. Option C is incorrect because reducing batch frequency impacts application responsiveness and does not address the I/O bottleneck.

156
MCQmedium

A company is running an Amazon Aurora MySQL-Compatible DB cluster. The application is experiencing increased latency during peak hours. The DB cluster's CPU utilization is at 40%, but the ReadReplicaLag metric is high. Which of the following is the most likely cause?

A.The writer instance is experiencing I/O bottlenecks, causing slow replication
B.The read replica is receiving too many write requests
C.The DB cluster parameter group is misconfigured
D.The binary log retention period is too long
AnswerA

I/O bottlenecks on the writer can delay writing to the binary log, increasing replica lag.

Why this answer

High replica lag indicates the replica is falling behind, often due to insufficient I/O capacity on the writer or replica. Option B is wrong because read replicas handle SELECT queries. Option C is wrong because the DB cluster parameter group is shared and misconfiguration is not the most likely cause for high replica lag.

Option D is wrong because binary log retention period does not affect replica lag directly.

157
MCQeasy

A company uses Amazon DynamoDB with auto scaling enabled. During a sales event, the write capacity consumption increases, but the table does not scale up as expected, resulting in throttled requests. The table has read/write capacity mode set to 'Provisioned' with auto scaling configured. What should the team check first to troubleshoot the issue?

A.Check the target utilization percentage in the auto scaling policy
B.Verify that the auto scaling role has the necessary IAM permissions
C.Check whether the table class is DynamoDB Standard-IA
D.Check if a global secondary index (GSI) has its own write capacity that is throttling
AnswerA

If target utilization is high (e.g., 90%), scaling may not trigger until sustained high usage.

Why this answer

The target utilization percentage in the auto scaling policy determines when scaling triggers. If set too high (e.g., 90%), the table may not scale until near full consumption, causing throttling during spikes. Option B is incorrect: IAM permissions for the auto scaling role are typically preconfigured and less likely to cause scaling failures—permission issues would appear in CloudTrail, not as a gradual throttling problem.

Option C is incorrect: table class (Standard vs. Standard-IA) affects storage costs, not write capacity scaling behavior. Option D is incorrect: while a GSI can throttle if its write capacity is insufficient, the primary table's auto scaling is independent; the question states the table itself does not scale, so the GSI is not the first check.

158
MCQmedium

A company is using Amazon ElastiCache for Redis as a caching layer. The application is experiencing higher latency than expected. The team suspects that cache evictions are occurring due to memory pressure. Which ElastiCache metric should be monitored to confirm this?

A.CurrConnections
B.Evictions
C.ReplicationLag
D.CacheHits
AnswerB

Evictions metric directly counts the number of keys evicted due to memory pressure.

Why this answer

The `Evictions` metric in ElastiCache for Redis tracks the number of keys evicted due to memory pressure, which directly indicates cache evictions. Option A is incorrect because `CurrConnections` shows the number of current client connections, not evictions. Option C is incorrect because `ReplicationLag` measures the delay between the primary and replica nodes, not evictions.

Option D is incorrect because `CacheHits` shows the number of successful key lookups, not evictions.

159
Multi-Selectmedium

A database specialist is monitoring an Amazon RDS for MySQL instance and notices that the 'CPUUtilization' metric is consistently above 90%. Which TWO actions should the specialist take to investigate the issue? (Select TWO.)

Select 2 answers
A.Disable the MySQL query cache to free up CPU.
B.Use Performance Insights to identify the queries that are consuming the most CPU.
C.Increase the instance class to provide more CPU resources.
D.Enable Multi-AZ to distribute the load to a standby instance.
E.Enable Enhanced Monitoring to get OS-level metrics for the RDS instance.
AnswersB, E

Performance Insights shows top queries by load.

Why this answer

Performance Insights identifies the queries that are consuming the most CPU, helping to locate the source of high CPU utilization. Option E is correct because Enhanced Monitoring provides OS-level metrics (e.g., CPU, memory, disk I/O) that help investigate the issue. Option A is incorrect because disabling the query cache does not help investigate CPU usage; it may even increase CPU load temporarily.

Option C is incorrect because increasing the instance class is a remediation step, not an investigation action. Option D is incorrect because Multi-AZ does not distribute read load to a standby instance; the standby is only for failover, not for load distribution.

160
MCQeasy

A company has an Amazon DynamoDB table with auto scaling enabled. During a traffic spike, the application experiences high write latencies. Which action should the company take to troubleshoot the latency issue?

A.Monitor the ThrottledWriteEvents metric in CloudWatch.
B.Switch the table to on-demand capacity mode.
C.Disable auto scaling and manually increase write capacity.
D.Increase the read capacity of the table.
AnswerA

Correct. Monitoring ThrottledWriteEvents in CloudWatch helps determine if auto scaling is keeping up with write demand. Elevated throttling indicates that provisioned capacity is insufficient, leading to high latencies.

Why this answer

ThrottledWriteEvents metric indicates write requests that were throttled due to exceeding provisioned capacity. If this metric spikes during traffic spikes, it suggests that auto scaling is not scaling fast enough, leading to high write latencies. Option B is incorrect because switching to on-demand is a remediation, not a troubleshooting step; it does not help identify the root cause.

Option C is incorrect because disabling auto scaling and manually increasing capacity might solve the issue temporarily but is not a troubleshooting approach; it bypasses auto scaling logic. Option D is incorrect because read capacity does not affect write latencies.

161
MCQhard

A company has an Amazon Aurora MySQL DB cluster with one writer and two readers. The application is reporting increased read latency. The DB cluster is using Aurora Auto Scaling for readers. Which metric should be used to trigger scaling out readers?

A.ReadLatency
B.ReaderNodeCPU
C.CPUUtilization
D.AverageActiveConnections
AnswerD

Aurora Auto Scaling uses the number of active connections to the reader to add or remove readers.

Why this answer

Aurora Auto Scaling for readers uses the 'AverageActiveConnections' metric to determine when to add readers. Option D is correct. Option A is incorrect because average active connections is the trigger, not read latency.

Option B is incorrect because ReaderNodeCPU is not a predefined metric for auto scaling. Option C is incorrect because CPUUtilization is not the metric used for reader scaling.

162
Multi-Selectmedium

Which TWO metrics should be monitored to detect an Amazon RDS for MySQL instance that is experiencing memory pressure? (Choose 2.)

Select 2 answers
A.SwapUsage
B.DatabaseConnections
C.ReadIOPS
D.FreeableMemory
E.CPUUtilization
AnswersA, D

High swap usage indicates memory pressure.

Why this answer

A and D are correct. SwapUsage directly indicates memory pressure when it exceeds 0, as the operating system swaps memory to disk. FreeableMemory shows the amount of available memory, so a low value suggests memory pressure.

B (DatabaseConnections) is incorrect because it is an indirect indicator that can affect memory but is not a direct memory metric. C (ReadIOPS) is disk I/O, not memory. E (CPUUtilization) is CPU usage, not memory.

163
Multi-Selecthard

Which THREE steps should be taken to troubleshoot high replica lag in an Amazon Aurora MySQL DB cluster? (Choose THREE.)

Select 3 answers
A.Review network latency between the writer and replica
B.Increase the instance size of the replica
C.Check the replica's CPU and memory utilization
D.Disable binary logging on the writer
E.Examine the replication thread status using SHOW SLAVE STATUS
AnswersA, C, E

Network latency can increase replication lag.

Why this answer

Options A, C, and E are correct. High replica lag in Amazon Aurora MySQL can be caused by network latency (A), resource contention on the replica (C), or issues with the replication process itself (E). Option B (increasing instance size) is a mitigation, not a troubleshooting step—you diagnose first.

Option D (disabling binary logging) would break replication entirely.

164
Multi-Selecteasy

An Amazon RDS for MySQL instance is running out of storage. Which TWO actions can be taken to resolve this issue without downtime?

Select 2 answers
A.Enable storage auto scaling to automatically add storage.
B.Delete unnecessary data from the instance.
C.Modify the DB instance to increase allocated storage.
D.Take a snapshot and restore to a larger instance.
E.Convert the instance to Amazon Aurora.
AnswersA, C

Auto scaling adds storage when needed without manual intervention.

Why this answer

Amazon RDS for MySQL supports modifying storage to increase allocated storage without downtime (option C). You can also enable storage auto scaling to automatically increase storage when thresholds are met (option A). Deleting unnecessary data (option B) might free space but is not a reliable proactive measure and could involve manual effort.

Taking a snapshot and restoring to a larger instance (option D) requires downtime. Converting to Amazon Aurora (option E) is a migration process, not a direct fix for running out of storage.

165
Multi-Selecteasy

A database administrator needs to monitor the free storage space on an Amazon RDS for SQL Server instance. Which TWO CloudWatch metrics should be used? (Choose two.)

Select 2 answers
A.DatabaseConnections
B.ReadLatency
C.BinLogDiskUsage
D.FreeStorageSpaceInPercent
E.FreeStorageSpace
AnswersD, E

Percentage of free storage space.

Why this answer

(FreeStorageSpaceInPercent) and Option E (FreeStorageSpace) are correct. Option A is wrong because DatabaseConnections measures the number of connections, not storage. Option B is wrong because ReadLatency measures read latency, not storage.

Option C is wrong because BinLogDiskUsage is a metric for MySQL/RDS for MySQL, not for SQL Server.

166
MCQhard

A company is running an Amazon DynamoDB table with on-demand capacity. The application is experiencing high latency for GetItem operations. CloudWatch shows that 'ThrottledRequests' is zero, but 'ConsumedReadCapacityUnits' is consistently below the provisioned limit. The table has a global secondary index (GSI) with a different partition key. What is the most likely cause of the high latency?

A.High write activity on the base table causing read latency
B.Insufficient provisioned read capacity on the base table
C.Throttling on the global secondary index due to uneven access patterns
D.Throttling on the base table due to hot partitions
AnswerD

Hot partitions on the base table can lead to increased latency even if ThrottledRequests is zero, as requests may queue internally before throttling occurs.

Why this answer

While throttling on a Global Secondary Index (GSI) can cause backpressure and affect writes to the base table, it does not directly impact read operations like GetItem. In this scenario, with zero ThrottledRequests, throttling is not occurring on the base table or GSI. The high latency is more likely caused by a hot partition on the base table, which can still result in increased response times even before throttling metrics are triggered.

Option C is therefore incorrect; the correct answer is D.

167
MCQhard

A database specialist is troubleshooting a degraded Amazon Aurora MySQL database cluster. The primary instance is showing elevated read latency and the replica lag is increasing. The DB cluster has one writer and one reader instance. The writer instance is using 90% of its allocated memory, and the reader instance is using 70%. Which action is most likely to reduce read latency and replica lag?

A.Increase the allocated storage for the cluster to improve I/O throughput.
B.Increase the writer instance size to provide more memory for write operations.
C.Increase the reader instance size to provide more buffer pool memory.
D.Add a second reader instance to distribute the read load.
AnswerC

More memory on the reader allows a larger buffer pool, reducing read latency and helping catch up replication.

Why this answer

Increasing the reader instance size provides more memory for the buffer pool, which improves read performance and can reduce replica lag. The reader is using 70% memory, so adding more buffer pool memory allows it to cache more data and handle read requests faster. Option A is incorrect because increasing storage improves I/O throughput but does not directly address memory constraints or replica lag.

Option B is incorrect because the writer instance is already under memory pressure, but the issue is read latency on the reader, not write performance. Option D is incorrect because adding a second reader distributes the read load but does not resolve the memory bottleneck on the existing reader; a larger instance is more effective in this scenario.

168
MCQhard

A company is using Amazon DynamoDB with auto scaling. The application is experiencing increased read latency. The monitoring shows that the ReadThrottleEvents metric is high. Which action should the company take to reduce read throttling?

A.Implement DynamoDB Accelerator (DAX) for caching.
B.Switch the table to on-demand capacity mode.
C.Enable DynamoDB Streams on the table.
D.Increase the write capacity units.
AnswerA

DAX caches reads, reducing pressure on the table.

Why this answer

DynamoDB Accelerator (DAX) provides an in-memory cache that reduces read latency and alleviates read throttling by serving frequently accessed items from cache, reducing the load on the underlying table. Option B is incorrect because switching to on-demand capacity mode may help with unpredictable traffic patterns but does not specifically address read throttling caused by high read demand; it can also increase costs. Option C is incorrect because enabling DynamoDB Streams does not directly reduce read throttling and may add overhead.

Option D is incorrect because increasing write capacity units does not affect read throttling since read and write capacities are separate.

169
MCQeasy

A developer reports that an Amazon RDS for PostgreSQL database instance's storage is full and the instance is in 'storage-full' state. The team wants to resolve this without downtime. Which action should be taken?

A.Create a read replica and promote it
B.Modify the DB instance to increase allocated storage
C.Delete old automated snapshots
D.Reboot the DB instance
AnswerB

RDS supports dynamic storage scaling without downtime for most engines.

Why this answer

Modifying the allocated storage for an RDS instance is a dynamic operation that can increase storage without downtime. Option A is wrong because creating a read replica requires the source to have available storage and does not resolve the storage issue on the primary instance. Option C is wrong because deleting old automated snapshots does not free up storage on the instance; snapshots are stored separately.

Option D is wrong because rebooting does not free storage; it only restarts the instance.

170
Multi-Selectmedium

A database administrator notices that an Amazon RDS for MySQL DB instance is experiencing high CPU utilization and increased latency during peak hours. The administrator wants to identify the queries causing the issue. Which TWO actions should be taken to diagnose the problem? (Select TWO.)

Select 2 answers
A.Enable Performance Insights on the DB instance.
B.Enable Amazon DevOps Guru for RDS.
C.Enable the slow query log and set a threshold for query duration.
D.Use RDS Proxy to manage connections.
E.Enable RDS Enhanced Monitoring.
AnswersA, C

Performance Insights provides detailed performance metrics and top SQL queries.

Why this answer

Enabling Performance Insights provides database performance metrics including wait events and SQL query details, allowing identification of queries causing high CPU and latency. Option C: Enabling the slow query log captures queries that exceed a specified duration threshold, helping to pinpoint slow-running queries. Option B (Amazon DevOps Guru for RDS) is an automated service for anomaly detection and performance insights, not a direct method to identify specific queries.

Option D (RDS Proxy) manages connection pooling and improves scalability, but does not identify problematic queries. Option E (RDS Enhanced Monitoring) provides OS-level metrics, not query-level details.

171
MCQhard

A company uses Amazon DynamoDB with auto scaling enabled. The application experiences increased latency during peak hours. The DynamoDB table has a read capacity of 10,000 RCU and write capacity of 5,000 WCU. The auto scaling target utilization is 70%. During peak hours, the consumed read capacity reaches 8,000 RCU, but auto scaling does not increase capacity. What is the most likely reason?

A.The consumed capacity is still below the provisioned capacity, so no throttling occurs, and auto scaling does not trigger.
B.The auto scaling configuration has a maximum capacity that prevents scaling beyond a certain limit.
C.Auto scaling for DynamoDB does not support increasing read capacity; it only decreases capacity.
D.Auto scaling only scales out when the consumed capacity exceeds the provisioned capacity.
AnswerB

If the maximum capacity is set to 10,000 RCU, auto scaling cannot increase further.

Why this answer

Auto scaling in DynamoDB adjusts capacity based on consumed capacity relative to provisioned capacity. With a target utilization of 70%, the expected consumed capacity for 10,000 RCU is 7,000 RCU. However, the actual consumed capacity is 8,000 RCU, which is 80% utilization, above the target.

Auto scaling should normally increase capacity to bring utilization back to 70%. But if the auto scaling configuration has a maximum read capacity limit (e.g., 8,000 RCU), scaling cannot exceed that limit, explaining why capacity does not increase. Option A is incorrect because auto scaling can trigger even before throttling occurs.

Option C is incorrect because auto scaling can increase both read and write capacity. Option D is incorrect because auto scaling scales based on sustained consumption above target utilization, not necessarily exceeding provisioned capacity.

172
MCQmedium

A company is using Amazon Redshift for data warehousing. The users report that queries are slow, and the CloudWatch metric 'CPUUtilization' shows high usage. The cluster has 4 dc2.large nodes. The administrator runs the following query: SELECT * FROM stl_query WHERE starttime > '2023-12-01' ORDER BY duration DESC LIMIT 10; This shows that the longest query runs for over 5 minutes. Which of the following is the MOST effective way to improve query performance?

A.Increase the size of the cluster volume by adding more storage
B.Add more nodes to the cluster (scale out)
C.Run the VACUUM command to reclaim space and improve query performance
D.Modify the workload management (WLM) queue to increase concurrency
AnswerB

Adding nodes increases compute capacity and parallelism.

Why this answer

Adding more nodes to the cluster (scale out) distributes the workload across additional compute resources, increasing parallelism and CPU capacity. This directly addresses the high CPU utilization and reduces query execution time. Option A is incorrect because the issue is CPU, not storage; increasing storage volume does not improve CPU performance.

Option C is incorrect: VACUUM reclaims disk space and sorts data but does not directly improve query performance when the bottleneck is CPU. Option D is incorrect: increasing WLM concurrency allows more queries to run simultaneously but may actually increase contention and not improve the performance of individual long-running queries.

173
MCQeasy

Refer to the exhibit. A database specialist is investigating an application outage that occurred at 14:25 UTC. The RDS event log shows the events above. What is the most likely cause of the outage?

A.A maintenance window update was applied, causing the instance to become unavailable.
B.The DB instance ran out of storage and became read-only.
C.A Multi-AZ failover occurred at 14:25 UTC, causing a brief outage.
D.The DB instance restarted at 14:30 UTC due to a manual reboot.
AnswerC

Failover involves promoting the standby, which causes a momentary outage.

Why this answer

The RDS event log shows 'Multi-AZ failover completed' at 14:25 UTC, which would cause a brief outage during the DNS switchover. The subsequent restart at 14:30 UTC is likely a result of the failover (e.g., instance recovery). Option A is wrong because there is no event indicating a maintenance window.

Option B is wrong because storage-full events would show before the outage, not a failover event. Option D is wrong because a manual reboot would not show a Multi-AZ failover event; the restart at 14:30 is after the failover.

174
Multi-Selecthard

A company's Amazon Redshift cluster is experiencing slow query performance. Which THREE diagnostic steps should be taken to identify the bottleneck?

Select 3 answers
A.Query STL_LOAD_ERRORS to check for data load errors
B.Query SVL_QUERY_REPORT to see step-level performance
C.Query STL_QUERY to analyze query execution times
D.Query STL_WLM_QUERY to check for queue waits
E.Query STV_BLOCKLIST to check disk allocation
AnswersB, C, D

Shows time spent per step.

Why this answer

To diagnose slow query performance in Amazon Redshift, you should query SVL_QUERY_REPORT to view step-level execution details (Option B), STL_QUERY to analyze query execution times (Option C), and STL_WLM_QUERY to check for WLM queue waits (Option D). Option A is incorrect because STL_LOAD_ERRORS is used for data load errors, not performance. Option E is incorrect because STV_BLOCKLIST shows disk block allocation, which is not directly related to query performance.

175
MCQmedium

A company is using Amazon ElastiCache for Redis as a caching layer for a web application. Users report that some cached data is missing, causing slower responses. Which ElastiCache feature should be checked first to understand key evictions?

A.ReplicationLag metric
B.CPUUtilization metric
C.Evictions metric in CloudWatch
D.CacheHits metric in CloudWatch
AnswerC

High evictions indicate memory pressure causing key removal.

Why this answer

The Evictions metric in CloudWatch directly shows the number of keys evicted from ElastiCache for Redis due to memory pressure. This is the first metric to check when cached data is missing, as evictions indicate that the cache is full and items are being removed to make space for new ones. Option A (ReplicationLag) measures replication delay between primary and replica nodes, not related to evictions.

Option B (CPUUtilization) shows CPU usage but does not indicate evictions. Option D (CacheHits) shows how often requested keys are found in the cache, but does not show evictions. Therefore, option C is correct.

176
MCQmedium

A company has an Amazon RDS for SQL Server database that stores customer orders. The database is running on a db.m5.large instance. Recently, the DBA noticed that the 'Write IOPS' metric is consistently high during business hours, and the 'Write Latency' metric is also elevated. The application is experiencing slow order placement. The DBA checks the disk queue depth and finds it to be high. The database uses General Purpose SSD (gp2) storage with 500 GB. What should the DBA do to improve write performance?

A.Enable Multi-AZ to offload writes to standby.
B.Change the storage type to gp3 and increase the IOPS.
C.Increase the storage size to 1000 GB to get more baseline IOPS.
D.Migrate to Provisioned IOPS (io2) storage with appropriate IOPS.
AnswerD

Provisioned IOPS provides consistent low latency for write-heavy workloads.

Why this answer

Provisioned IOPS (io2) storage provides consistent, high IOPS performance with low latency, which is needed when the workload exceeds the baseline IOPS of gp2 storage. The current gp2 volume at 500 GB provides only 1,500 baseline IOPS (3 IOPS per GB), which is insufficient for the high write IOPS during business hours, causing high write latency and disk queue depth. Option A is incorrect because Multi-AZ does not offload writes; it only provides high availability and may increase write latency due to synchronous replication.

Option B is incorrect because while gp3 offers baseline performance with the ability to provision additional IOPS independently of storage size, it is still a general-purpose SSD; for consistent low latency under sustained high IOPS, io2 is more suitable. Option C is incorrect because increasing gp2 storage to 1,000 GB would increase baseline IOPS to 3,000, but this may not be enough, and it is more cost-effective to use io2 with the required IOPS rather than over-provisioning storage.

177
MCQhard

An e-commerce application uses Amazon Aurora MySQL. During a flash sale, the database experiences a surge in write operations, and the application reports 'Lock wait timeout exceeded' errors. The DBA suspects a long-running transaction is blocking others. Which AWS tool or feature should the DBA use to identify the blocking transaction?

A.Amazon Aurora Performance Insights with the 'Wait for lock' dimension
B.Amazon Inspector
C.CloudWatch Logs Insights
D.Amazon RDS Enhanced Monitoring
AnswerA

Performance Insights can display blocking sessions and wait events.

Why this answer

Amazon Aurora Performance Insights with the 'Wait for lock' dimension can identify blocking sessions by showing which sessions are waiting on locks and which session holds the lock. Option B (Amazon Inspector) is a security assessment tool and does not provide database lock information. Option C (CloudWatch Logs Insights) can analyze logs but does not directly identify blocking transactions.

Option D (Amazon RDS Enhanced Monitoring) provides OS-level metrics, not database-level lock information.

178
MCQmedium

A company is using Amazon RDS for SQL Server with Multi-AZ. The application experiences a brief interruption during a planned maintenance window. The DBA wants to verify that the failover occurred correctly. Which set of metrics should the DBA examine?

A.FreeStorageSpace on the primary instance
B.DatabaseConnections and ReplicaLag
C.WriteIOPS and ReadIOPS for both instances
D.NetworkThroughput on the replica instance
AnswerB

Connections drop to zero during failover, and ReplicaLag becomes zero after failover.

Why this answer

During a Multi-AZ failover, DatabaseConnections drop to zero as existing connections are terminated, then rise again as applications reconnect. ReplicaLag, which measures the delay between primary and standby, becomes zero after failover because the former standby is now the primary. Monitoring these two metrics together provides clear evidence of a failover event.

Option A (FreeStorageSpace) does not change during failover. Option C (WriteIOPS and ReadIOPS) may fluctuate but are not direct indicators of failover. Option D (NetworkThroughput) may also vary but is not a specific failover indicator.

179
Multi-Selectmedium

A company is using Amazon Redshift and has a query that is running slowly. The DBA wants to identify if the query is I/O-bound. Which TWO metrics from Amazon CloudWatch can indicate I/O-bound queries?

Select 2 answers
A.NetworkReceiveThroughput
B.ReadIOPS
C.CPUUtilization
D.WriteIOPS
E.ReadLatency
AnswersB, E

High ReadIOPS indicates many I/O operations.

Why this answer

Options B and E are correct. ReadIOPS measures the number of read I/O operations per second, and ReadLatency measures the average time per read I/O operation. High values for these metrics indicate that the query is I/O-bound.

Option A (NetworkReceiveThroughput) measures network throughput, not I/O. Option C (CPUUtilization) indicates CPU-bound workloads. Option D (WriteIOPS) measures write I/O, which is less likely to be the primary indicator for a slow query, and the question asks for the two metrics that can indicate I/O-bound queries; ReadIOPS and ReadLatency are the most direct I/O metrics.

180
Multi-Selectmedium

A company is using Amazon DynamoDB for a session management application. The application is experiencing higher latency during peak hours. The DynamoDB table uses on-demand capacity. Which TWO steps should a database specialist take to diagnose the issue?

Select 2 answers
A.Review the table's ReadCapacityUnits and WriteCapacityUnits to ensure they are sufficient.
B.Examine the CloudWatch metric for UserErrors to identify client-side issues.
C.Check the CloudWatch metric for ThrottledRequests to see if any requests are being throttled.
D.Enable DynamoDB Streams to capture all write events for analysis.
E.Analyze the PartitionKey metrics to detect if a single partition is receiving a disproportionate amount of traffic.
AnswersC, E

Throttling can cause increased latency.

Why this answer

Checking ThrottledRequests helps identify if requests are being throttled, even with on-demand. Reviewing PartitionKey metrics helps detect hot partitions. ReadCapacityUnits and WriteCapacityUnits are not applicable for on-demand.

Latency metrics show end-to-end performance but are less diagnostic.

181
MCQhard

Refer to the exhibit. A database specialist is troubleshooting an automation script that fails when trying to create a snapshot of the RDS DB instance 'mydb' using an IAM role with the attached policy. The error message indicates that the user is not authorized to perform the operation. Which statement best explains the failure?

A.The resource ARN for the snapshot is incorrect; it should specify the DB instance ID.
B.The policy does not allow the rds:DescribeDBInstances action on the snapshot resource.
C.The policy does not allow the rds:CreateDBSnapshot action on the specific snapshot name.
D.The policy does not grant the rds:CreateDBSnapshot permission on the DB instance resource.
AnswerD

CreateDBSnapshot requires permission on the DB instance.

Why this answer

The rds:CreateDBSnapshot action requires permission on the DB instance resource (arn:aws:rds:region:account:db:instance-name). The IAM policy in the exhibit only grants this action on the snapshot resource (arn:aws:rds:region:account:snapshot:*), not on the DB instance, causing the authorization failure. Options A and B are incorrect because they either mention an irrelevant action or misidentify the resource.

Option C is incorrect because the policy does allow rds:CreateDBSnapshot on the snapshot resource, but the missing permission on the DB instance is the actual issue.

182
MCQeasy

An administrator is troubleshooting an Amazon RDS for PostgreSQL instance that is experiencing high CPU utilization. The administrator has enabled Performance Insights. Which metric should be examined first to identify the queries consuming the most CPU?

A.db.sessions
B.db.cpu.avg
C.db.load.avg
D.db.bytes_sent
AnswerC

This metric shows the average number of active sessions and is key for identifying high-load queries.

Why this answer

The correct metric to examine first is 'db.load.avg' (Option C). In Amazon RDS Performance Insights, 'db.load.avg' represents the average number of active sessions, which directly correlates with CPU utilization. A high value indicates queries that are consuming significant CPU resources.

Option A ('db.sessions') is not a standard Performance Insights metric. Option B ('db.cpu.avg') does not exist in Performance Insights. Option D ('db.bytes_sent') is a network metric unrelated to CPU usage.

183
Multi-Selectmedium

A company is experiencing slow query performance on an Amazon RDS for MySQL database. The DBA wants to identify the most time-consuming queries. Which TWO actions should the DBA take? (Choose two.)

Select 2 answers
A.Enable the audit log.
B.Enable the general log and review it.
C.Use Amazon RDS Enhanced Monitoring.
D.Use Amazon RDS Performance Insights.
E.Enable the slow query log and monitor it in CloudWatch Logs.
AnswersD, E

Performance Insights identifies top SQL by load.

Why this answer

Amazon RDS Performance Insights (Option D) provides a database performance tuning and monitoring feature that visualizes database load and identifies the most time-consuming queries by breaking down wait events, SQL statements, and hosts. It directly helps the DBA pinpoint the specific queries causing performance degradation without additional configuration or overhead.

Exam trap

The trap here is that candidates often confuse general logging (Option B) with slow query logging, or assume Enhanced Monitoring (Option C) provides query-level insights, when in fact only Performance Insights and the slow query log directly identify the most time-consuming queries.

184
Multi-Selecthard

A company uses Amazon Aurora MySQL for its e-commerce platform. The DB cluster has one writer and two readers. Recently, the application started showing occasional deadlock errors during order processing. The error logs show: 'Transaction (Process ID 123) was deadlocked on lock resources with another process and has been chosen as the deadlock victim. Rerun the transaction.' The application retries three times before failing. The development team wants to reduce the likelihood of deadlocks. Which three actions should the team take? (Choose three.)

Select 3 answers
A.Shorten the duration of transactions by committing frequently.
B.Switch to READ UNCOMMITTED isolation level to reduce locking.
C.Use READ COMMITTED isolation level instead of REPEATABLE READ.
D.Increase the number of retry attempts to 10.
E.Ensure that transactions access tables in the same order.
AnswersA, C, E

Shorter transactions hold locks for less time, reducing the chance of conflicts.

Why this answer

Reduces the time locks are held, decreasing contention. Option C uses a weaker isolation level (READ COMMITTED) that reduces locking overhead compared to REPEATABLE READ. Option E ensures transactions request locks in a consistent order, preventing cycle dependencies.

Option B is incorrect because READ UNCOMMITTED can cause dirty reads and is not suitable for e-commerce order processing. Option D is incorrect because increasing retry attempts does not reduce the likelihood of deadlocks; it only allows more retries after a deadlock occurs.

185
Multi-Selecteasy

Which TWO CloudWatch metrics should be monitored to determine if an Amazon Aurora MySQL DB cluster has sufficient I/O capacity?

Select 2 answers
A.ReadIOPS
B.WriteIOPS
C.FreeableMemory
D.CPUUtilization
E.DatabaseConnections
AnswersA, B

ReadIOPS shows the number of read I/O operations per second.

Why this answer

The correct metrics to monitor for sufficient I/O capacity are ReadIOPS (Option A) and WriteIOPS (Option B). These metrics represent the actual number of read and write I/O operations per second. If the cluster is approaching the maximum IOPS limit of the underlying storage, increasing I/O capacity may be needed.

Option C (FreeableMemory) is related to memory, not I/O. Option D (CPUUtilization) measures CPU usage, not I/O. Option E (DatabaseConnections) tracks connections, not I/O capacity.

186
MCQhard

A company runs an Amazon Aurora MySQL database cluster with one writer and two readers. The application suddenly fails with 'Too many connections' error. The writer instance's maximum connections is set to 1000. Which configuration change would best resolve the issue while maintaining high availability?

A.Enable Amazon RDS Proxy for the database cluster.
B.Increase the max_connections parameter on the writer instance to 5000.
C.Redirect all write requests to one of the read replicas.
D.Manually kill idle connections from the database.
AnswerA

Correct. RDS Proxy efficiently manages database connections, reducing the number of open connections.

Why this answer

Amazon RDS Proxy acts as a connection pooler that manages database connections efficiently. By reducing connection churn and reusing connections, it prevents the 'Too many connections' error without overloading the instance. Option B is incorrect because simply increasing max_connections could lead to resource exhaustion and degradation of performance.

Option C is incorrect because read replicas cannot accept write traffic; they are read-only. Option D is incorrect because manually killing idle connections is a temporary fix and not a configuration change; it does not prevent the issue from recurring.

187
MCQmedium

A company is experiencing increased latency on their RDS for PostgreSQL instance. The CloudWatch metrics show high ReadIOPS but low CPU utilization. Which action is MOST likely to resolve the issue?

A.Migrate the database to Amazon Aurora PostgreSQL
B.Increase the instance size or switch to a gp3 volume
C.Enable Multi-AZ deployment
D.Enable storage auto-scaling
AnswerB

More memory and I/O capacity reduces wait events.

Why this answer

High ReadIOPS with low CPU indicates an I/O bottleneck. Increasing the instance size provides more dedicated IOPS, and switching to gp3 volume offers better baseline performance and throughput compared to gp2. Option A is incorrect because migrating to Aurora does not directly address the I/O bottleneck and involves additional considerations.

Option C is incorrect because Multi-AZ provides high availability, not improved performance. Option D is incorrect because storage auto-scaling only increases storage capacity, not IOPS or throughput.

188
MCQeasy

A developer is troubleshooting an issue where an application using Amazon DynamoDB is receiving occasional 'ThrottlingException' errors. The application uses eventually consistent reads. What is the MOST likely cause of this error?

A.The application is using an incorrect table name
B.The read capacity units are set too low for the current traffic pattern
C.The application is using eventually consistent reads instead of strongly consistent reads
D.There is a network connectivity issue between the application and DynamoDB
AnswerB

Throttling happens when traffic exceeds provisioned capacity.

Why this answer

ThrottlingException occurs when the request rate exceeds the provisioned throughput capacity. Option A is incorrect because an incorrect table name would result in a ResourceNotFoundException, not a ThrottlingException. Option C is incorrect because eventually consistent reads consume half the read capacity units compared to strongly consistent reads, making them less likely to cause throttling.

Option D is incorrect because network connectivity issues would typically cause timeout or connection errors, not ThrottlingException.

189
Multi-Selecteasy

A company uses Amazon DynamoDB Global Tables for a multi-region application. The table is configured with on-demand capacity. The application writes data in the us-east-1 region and reads from us-west-2. Users in us-west-2 report that data written in us-east-1 is not appearing in us-west-2 within the expected replication latency of under 5 seconds. Instead, replication sometimes takes up to 30 seconds. Which two factors could be causing this increased replication latency? (Choose two.)

Select 2 answers
A.The application is reading from the us-west-2 table before replication completes.
B.High network latency or packet loss between us-east-1 and us-west-2.
C.The read capacity in us-west-2 is insufficient, causing read throttling.
D.The on-demand capacity mode is causing write throttling in us-east-1.
E.A large volume of writes to the table creates a backlog in the replication stream.
AnswersB, E

Replication relies on network connectivity; latency increases replication time.

Why this answer

Options B and E are correct. Option B: High network latency or packet loss between the two regions directly increases replication time. Option E: A large volume of writes creates a backlog in the Global Tables replication stream, causing delays.

Option A is incorrect because reading before replication completes is a symptom, not a cause of increased latency. Option C is incorrect because read capacity in a replica region does not affect replication latency; replication uses write capacity. Option D is incorrect because on-demand capacity automatically scales and does not cause write throttling.

190
MCQmedium

A company is using Amazon Neptune and notices that some queries are slow. The DBA wants to identify which queries consume the most time. Which feature should be used?

A.Database audit log
B.Query profiler
C.Slow query log
D.Performance Insights
AnswerB

Query profiler captures execution details of queries.

Why this answer

Neptune's query profiler provides detailed information about query execution time. The slow query log only logs queries that exceed a threshold. The audit log is for security events.

The performance insights is for RDS, not Neptune.

191
MCQmedium

A company uses Amazon ElastiCache for Redis as a caching layer for a web application. They observe a sudden increase in CPU utilization on the cache cluster, and the application experiences higher latency. Which action should be taken to diagnose the issue?

A.Enable query profiling on the ElastiCache cluster.
B.Monitor Amazon CloudWatch metrics for the ElastiCache cluster to identify the cause.
C.Disable persistence to reduce CPU overhead.
D.Immediately scale up to a larger node type.
AnswerB

Correct. CloudWatch metrics provide insight into CPU usage, cache hits/misses, and other performance indicators.

Why this answer

The first step in diagnosing a sudden CPU utilization increase and latency in ElastiCache is to monitor CloudWatch metrics such as CPUUtilization, CacheHits, CacheMisses, and EngineCPUUtilization to identify patterns or bottlenecks. Option A is incorrect because ElastiCache for Redis does not support query profiling; that feature is for RDS. Option C is incorrect because disabling persistence (e.g., AOF or RDB) reduces durability but is not a diagnostic action and may not address CPU issues.

Option D is incorrect because scaling up should only be considered after analyzing metrics to determine if capacity is the actual problem.

192
MCQhard

A company's DynamoDB table uses provisioned capacity and has a global table for disaster recovery. Write requests to the replica region are failing with ProvisionedThroughputExceededException even though the replica table shows low consumed capacity. What is the MOST likely cause?

A.Auto-scaling is disabled on the replica table
B.The replica table is exhausting its burst capacity
C.The replica table's write capacity is insufficient for replication writes
D.DAX is enabled on the replica table
AnswerC

The replica table's write capacity is insufficient for replication writes. This is correct because global table replication consumes write capacity on the replica, and if that capacity is insufficient, writes are throttled.

Why this answer

In DynamoDB global tables, writes to the replica region consume write capacity units (WCUs) on the replica table. If the provisioned write capacity on the replica table is insufficient to handle the replication traffic, the replica will throttle incoming write requests with a ProvisionedThroughputExceededException, even if the consumed capacity appears low because throttling occurs before the write is counted. Option C correctly identifies this cause.

Option A is incorrect because auto-scaling adjusts capacity over time but does not directly cause throttling. Option B is incorrect because burst capacity provides short-term flexibility; sustained throttling is due to insufficient capacity, not burst exhaustion. Option D is incorrect because DAX is an in-memory cache for reads and does not affect write operations.

193
MCQmedium

A company is using Amazon Aurora MySQL. The database performance has degraded, and the DBA wants to identify the queries that are waiting for locks. Which system table should be queried to find information about locking conflicts?

A.information_schema.TABLES
B.information_schema.PROCESSLIST
C.performance_schema.events_waits_current
D.information_schema.INNODB_LOCKS
AnswerD

INNODB_LOCKS shows current InnoDB locks.

Why this answer

Information_schema.INNODB_LOCKS. This table shows current locks held by InnoDB transactions, including lock type, mode, and which transactions are waiting for locks. Option A (information_schema.TABLES) provides metadata about tables, not locks.

Option B (information_schema.PROCESSLIST) shows current running processes but not lock details. Option C (performance_schema.events_waits_current) shows current wait events but is not specific to InnoDB lock conflicts; INNODB_LOCKS is the correct source for identifying locking conflicts in Aurora MySQL.

194
MCQeasy

A database administrator notices that an Amazon RDS for Oracle instance has a high number of connections, causing performance degradation. Which tool can be used to identify the active sessions and their queries?

A.Amazon Inspector
B.Performance Insights
C.RDS Enhanced Monitoring
D.AWS CloudTrail
AnswerB

Performance Insights shows active sessions and their queries.

Why this answer

Performance Insights provides a dashboard to monitor active sessions and the queries they are running. Option A is incorrect because Amazon Inspector is a security assessment tool, not a monitoring tool for database sessions. Option C is incorrect because RDS Enhanced Monitoring provides OS-level metrics like CPU and memory, not database session details.

Option D is incorrect because AWS CloudTrail logs API calls, not database sessions or queries.

195
MCQeasy

A database administrator notices that the /var/log/mysql/error.log file on an Amazon RDS for MySQL DB instance is growing rapidly. The administrator wants to monitor the log file size and receive alerts when it exceeds a certain threshold. Which AWS service should be used to set up this monitoring?

A.AWS CloudTrail
B.Amazon S3
C.Amazon RDS Enhanced Monitoring
D.Amazon CloudWatch Logs
AnswerD

CloudWatch Logs can monitor log file size and trigger alarms.

Why this answer

CloudWatch Logs can monitor log file sizes and trigger alarms based on metric filters. Option A is wrong because CloudTrail does not monitor log file sizes. Option B is wrong because S3 is a storage service, not a monitoring service.

Option C is wrong because RDS Enhanced Monitoring provides OS-level metrics, not log file metrics.

196
Multi-Selecthard

A company is using Amazon DynamoDB with global tables. The application is experiencing high write latency in one region. Which THREE factors could contribute to this issue? (Choose three.)

Select 3 answers
A.Replication lag from global tables causing write conflicts.
B.Insufficient read capacity units provisioned for the table.
C.High network latency between the application and the DynamoDB endpoint.
D.The table's auto scaling configuration is set to increase write capacity too aggressively.
E.Insufficient write capacity units provisioned for the table.
AnswersA, C, E

Global tables replicate asynchronously, and conflicts can increase latency.

Why this answer

Replication lag in global tables can lead to write conflicts and increased latency. Option C is correct because high network latency between the application and the DynamoDB endpoint directly impacts write latency. Option E is correct because insufficient write capacity units provisioned can cause throttling, resulting in higher write latency.

Option B is incorrect because read capacity units do not affect write throughput. Option D is incorrect because aggressive auto scaling increases capacity, which would reduce latency, not increase it.

197
Multi-Selecteasy

A company is troubleshooting an Amazon DynamoDB table that is experiencing high latency. The table uses on-demand capacity. Which TWO steps should be taken to diagnose the issue?

Select 2 answers
A.Increase the provisioned read capacity
B.Examine CloudWatch metrics for throttling and latency patterns
C.Consider using DynamoDB Accelerator (DAX) for caching
D.Disable TTL to reduce write overhead
E.Split the table into multiple partitions manually
AnswersB, C

CloudWatch metrics provide insights into throttling, latency, and consumed capacity, which are essential for diagnosing performance issues.

Why this answer

CloudWatch metrics for DynamoDB can show throttling events, latency patterns, and consumed throughput, which help diagnose high latency issues. Option C is correct because DynamoDB Accelerator (DAX) provides an in-memory cache that can significantly reduce read latency for frequently accessed items. Option A is incorrect because the table uses on-demand capacity, so adjusting provisioned capacity is not applicable.

Option D is incorrect because disabling TTL does not directly reduce write overhead or latency. Option E is incorrect because DynamoDB handles partitions automatically and manual splitting is not supported.

198
MCQmedium

Refer to the exhibit. A database specialist is investigating performance degradation on an Amazon RDS for MySQL DB instance. The BurstBalance metric shows the values above. What does this indicate, and what action should be taken?

A.The instance has exhausted its I/O burst credits; scale up the allocated storage to increase baseline I/O.
B.The instance is experiencing high read load; add a read replica to offload reads.
C.The instance's compute capacity is insufficient; change the instance class to a larger size.
D.The instance is experiencing a memory bottleneck; enable Performance Insights to analyze query performance.
AnswerA

Larger storage volumes have higher baseline I/O and accumulate burst credits faster.

Why this answer

The BurstBalance metric dropping to 0 indicates that the instance has exhausted its I/O burst credits, resulting in I/O throttling. Scaling up the allocated storage increases the baseline I/O performance, which in turn increases the burst credit earning rate and raises the burst balance. Option B is incorrect because adding a read replica offloads read traffic but does not increase I/O credits.

Option C is incorrect because changing the instance class does not affect I/O credits; it only increases compute capacity. Option D is incorrect because enabling Performance Insights helps with diagnosis but does not resolve I/O credit exhaustion.

199
MCQmedium

Refer to the exhibit. A DBA sees the above error log entries for an Amazon RDS for MySQL DB instance. Which action should the DBA take to resolve the 'Too many connections' error?

A.Modify the DB parameter group to increase max_connections
B.Reboot the DB instance
C.Modify the DB security group to allow more connections
D.Reset the DB instance master user password
AnswerA

Increasing max_connections allows more connections.

Why this answer

Increasing the max_connections parameter in the DB parameter group allows the MySQL instance to handle more concurrent connections, directly resolving the 'Too many connections' error. Option B is incorrect because rebooting the DB instance does not change the connection limit, only temporarily clears existing connections. Option C is incorrect because modifying the DB security group adjusts network access rules, not the database's internal connection limit.

Option D is incorrect because resetting the master user password addresses authentication issues, not the connection limit.

200
MCQmedium

A database engineer is monitoring an Amazon RDS for PostgreSQL instance and notices that the 'DiskQueueDepth' metric is consistently above 100. The instance uses gp2 storage with 1000 GB allocated. What is the most likely cause of the high disk queue depth?

A.Replication lag between the primary and standby instance
B.The instance has reached the IOPS limit of the gp2 volume
C.Insufficient memory allocated to the instance
D.Network throughput limit is being exceeded
AnswerB

For a 1000 GB gp2 volume, baseline IOPS is 3000; sustained I/O beyond that causes queuing.

Why this answer

A consistently high DiskQueueDepth (above 100) on an Amazon RDS for PostgreSQL instance with gp2 storage indicates that the volume is saturating its IOPS limit. gp2 volumes provide a baseline of 3 IOPS per GB (up to 16,000 IOPS), so a 1000 GB gp2 volume has a baseline of 3000 IOPS. When the workload exceeds this baseline, the volume relies on burst credits, and once credits are exhausted, IOPS are throttled to the baseline, causing I/O requests to queue up and the DiskQueueDepth metric to rise.

Exam trap

The trap here is that candidates may confuse DiskQueueDepth with memory or network metrics, or assume that any high queue depth automatically indicates a hardware failure, rather than recognizing it as a symptom of IOPS exhaustion on gp2 storage.

How to eliminate wrong answers

Option A is wrong because replication lag between primary and standby instances is measured by the 'ReplicaLag' metric, not DiskQueueDepth, and it does not directly cause high disk queue depth on the primary instance. Option C is wrong because insufficient memory allocated to the instance would manifest as high swap usage or low FreeableMemory, not as a high DiskQueueDepth, which is a storage I/O metric. Option D is wrong because network throughput limits are tracked by metrics like 'NetworkThroughput' or 'NetworkPacketsIn/Out', and exceeding them would cause packet loss or latency, not a buildup of I/O requests at the disk level.

201
MCQeasy

A database administrator notices that the CPU utilization on an Amazon RDS for PostgreSQL instance is consistently above 90% during peak hours. Which CloudWatch metric should be checked first to identify the cause of the high CPU usage?

A.DatabaseConnections
B.NetworkThroughput
C.SwapUsage
D.ReadIOPS
AnswerA

High connections can lead to high CPU from session management.

Why this answer

A high number of database connections can lead to increased CPU usage as each connection requires processing. Option B is wrong because NetworkThroughput measures network traffic, which is not a primary cause of high CPU. Option C is wrong because SwapUsage indicates memory pressure, not CPU.

Option D is wrong because ReadIOPS measures disk I/O and may not directly cause high CPU.

202
MCQmedium

A company's Amazon RDS for PostgreSQL instance is running out of storage. The DB instance has auto-scaling enabled, but the storage did not increase. What is the most likely cause?

A.The DB instance class does not support storage auto-scaling.
B.The DB instance is smaller than the minimum storage size for auto-scaling.
C.The DB instance has exceeded the Maximum Storage Duration setting.
D.The storage usage has not reached the maximum allocated storage.
AnswerC

As the most likely cause is that the maximum storage threshold has been reached, preventing further auto-scaling.

Why this answer

Amazon RDS storage auto-scaling will not increase storage beyond the maximum storage threshold set by the user. If the instance has reached this maximum threshold, auto-scaling stops, even if free space is low. The phrase 'Maximum Storage Duration' in the option refers to this threshold.

Option D is incorrect because the condition for auto-scaling is based on free space being below 10%, not on whether the maximum storage has been reached. If the maximum threshold has not been reached, auto-scaling should occur when free space is low.

Exam trap

Candidates often confuse 'maximum storage threshold' with 'allocated storage'. The threshold is the upper limit for auto-scaling; once reached, no further scaling occurs.

203
MCQeasy

A company runs an Amazon RDS for Oracle DB instance. The database administrator wants to receive an alert when the storage space is below 10% of the allocated storage. Which Amazon CloudWatch metric and alarm threshold should be used?

A.Metric: FreeStorageSpace, Condition: < 10% of 100 GB (10 GB)
B.Metric: BinaryLogUsage, Condition: > 10%
C.Metric: FreeableMemory, Condition: < 10% of total memory
D.Metric: DiskQueueDepth, Condition: > 10
AnswerA

FreeStorageSpace metric with a threshold of 10 GB (assuming 100 GB allocated) would trigger when free space is below 10%.

Why this answer

The correct metric is FreeStorageSpace, which reports the available storage in bytes. To alert when storage is below 10% of allocated, compute the threshold as 10% of the total allocated storage (e.g., 10 GB for 100 GB). Option A defines the condition correctly.

Option B is wrong because BinaryLogUsage applies to MySQL, not Oracle. Option C is incorrect as FreeableMemory tracks memory, not storage. Option D is wrong because DiskQueueDepth measures I/O queue depth, not storage capacity.

204
MCQmedium

A database administrator is troubleshooting an Amazon RDS for SQL Server instance that is experiencing high 'ReadIOPS' and 'ReadLatency'. The instance uses General Purpose SSD (gp2) storage. The 'BurstBalance' metric is 0%. What should the administrator do to improve performance?

A.Enable Multi-AZ to distribute the load
B.Increase the allocated storage or switch to Provisioned IOPS
C.Create a read replica to offload read traffic
D.Disable automatic backups to reduce I/O
AnswerB

Increasing volume size increases baseline IOPS for gp2; switching to io1/io2 provides consistent IOPS.

Why this answer

When BurstBalance is 0%, the gp2 volume has exhausted its burst credits and is operating at baseline IOPS. To improve performance, you can increase the volume size (which increases baseline IOPS) or switch to Provisioned IOPS (io1/io2) for consistent performance. Option A is wrong because enabling Multi-AZ does not increase I/O performance; it provides high availability.

Option C is wrong because creating a read replica offloads read traffic but does not improve write performance or reduce latency on the primary instance. Option D is wrong because disabling automatic backups reduces storage I/O but does not directly improve ReadIOPS or ReadLatency; the primary issue is exhausted burst credits.

205
MCQeasy

A developer executed a DELETE statement without a WHERE clause on an Amazon RDS for PostgreSQL instance. The transaction is still open. Which action should the developer take to undo the DELETE without affecting other operations?

A.Execute COMMIT and then run a recovery script.
B.Use the Point-in-Time Recovery feature to restore the database to a time before the DELETE.
C.Execute ROLLBACK in the same session.
D.Stop the DB instance and restore from the latest snapshot.
AnswerC

ROLLBACK undoes all changes made in the current transaction.

Why this answer

The correct action is to execute ROLLBACK in the same session. Since the transaction is still open, issuing ROLLBACK will undo the DELETE statement without affecting other operations. Option A is incorrect because COMMIT would make the DELETE permanent.

Option B is incorrect because Point-in-Time Recovery restores the entire DB instance to a past state, which is unnecessary and affects other data. Option D is incorrect because stopping and restoring from a snapshot would also affect other data and is not the simplest solution.

206
MCQhard

A company is running an Amazon DynamoDB table with provisioned capacity. The table has a partition key of 'user_id' and a sort key of 'timestamp'. The application performs frequent Query operations using the partition key and a range of sort keys. Recently, the 'ThrottledRequests' metric has spiked. The read and write capacity units are consistently at 80% utilization. What is the most effective way to resolve the throttling?

A.Increase the provisioned read and write capacity units
B.Add a global secondary index with a different partition key
C.Change the partition key to a more uniformly distributed attribute
D.Enable DynamoDB Auto Scaling with a higher target utilization
AnswerC

Correct: Redesigning the partition key to be more uniform evenly distributes the workload, directly addressing the root cause of throttling.

Why this answer

The throttling is likely due to uneven access patterns creating hot partitions. By changing the partition key to a more uniformly distributed attribute, the load is spread evenly across partitions, reducing throttling without needing to increase capacity. Option D is incorrect because enabling Auto Scaling with a higher target utilization delays scaling and can worsen throttling on hot partitions; it does not address the root cause.

Option A is incorrect because increasing capacity may not help if a hot partition is throttled, and it increases costs. Option B is incorrect because adding a GSI does not reduce throttling on the base table; it only provides an alternative access pattern.

Exam trap

Auto Scaling with a higher target utilization seems like a quick fix but actually makes hot partition problems worse by allowing higher utilization before scaling.

207
MCQhard

A company is using Amazon DynamoDB Accelerator (DAX) to improve read performance. Recently, the cache hit ratio has dropped significantly. The application uses strongly consistent reads. What is the most likely cause of the low cache hit ratio?

A.The DynamoDB table's write capacity is too low
B.The application is using strongly consistent reads, which bypass the DAX cache
C.The DAX cluster has too few nodes
D.The DAX cluster's TTL is set too low
AnswerB

DAX only caches eventually consistent reads; strongly consistent reads go directly to DynamoDB.

Why this answer

DAX (DynamoDB Accelerator) is designed to cache data for eventually consistent reads only. When an application uses strongly consistent reads, those requests bypass the DAX cache entirely and go directly to DynamoDB, resulting in a low cache hit ratio. Option A is incorrect because low write capacity may cause throttling but does not directly affect the cache hit ratio.

Option C is incorrect because having too few nodes can impact performance and availability, but the primary cause of the drop in cache hits is the consistency model mismatch. Option D is incorrect because while a low TTL can reduce cache effectiveness, the key issue here is that strongly consistent reads never use the cache.

208
MCQmedium

A company is running a production Amazon RDS for MySQL DB instance. The application team reports intermittent connection timeouts. The DBA notices that the DB instance's CPU utilization spikes to 100% during these times. Which metric should be monitored to determine if the issue is due to a specific query?

A.DatabaseConnections
B.NetworkThroughput
C.ReadIOPS
D.Queries (engine-specific counter)
AnswerD

Queries reflects the number of queries executed, helping identify query load spikes.

Why this answer

The RDS for MySQL engine-specific counter 'Queries' reflects the number of queries executed. Option A is wrong because DatabaseConnections shows connections, not query performance. Option B is wrong because ReadIOPS measures disk I/O, not query volume.

Option C is wrong because NetworkThroughput measures network traffic.

209
MCQhard

A company uses Amazon DynamoDB Global Tables with strong consistent reads. They notice that a write to us-east-1 is not visible in eu-west-1 after several seconds. Which configuration setting is MOST likely causing this behavior?

A.DynamoDB Streams is not enabled on the table
B.Auto Scaling is configured for write capacity
C.Last writer wins (LWW) conflict resolution is disabled
D.Strongly consistent reads are used on a global table
AnswerA

Global Tables require DynamoDB Streams to replicate writes; without it, replication does not occur.

Why this answer

DynamoDB Global Tables require DynamoDB Streams to be enabled on the table. If streams are disabled, write operations in one region will not be replicated to other regions, causing the writes to never become visible. This matches the symptom of writes not appearing after several seconds.

Option B (Auto Scaling) does not affect replication. Option C (LWW conflict resolution) is enabled by default and does not cause delays. Option D (strongly consistent reads) are not supported on Global Tables and would generate an error, not a delay.

210
Multi-Selecthard

Which THREE metrics should be monitored in Amazon CloudWatch to detect a potential memory leak in an Amazon RDS for SQL Server instance? (Choose three.)

Select 3 answers
A.DatabaseConnections
B.CPUUtilization
C.ReadIOPS
D.SwapUsage
E.FreeableMemory
AnswersA, D, E

If connections are not released, memory usage may increase.

Why this answer

Options A, D, and E are correct. DatabaseConnections can indicate a memory leak if connections are not closed, consuming memory. SwapUsage indicates memory pressure when physical memory is insufficient.

FreeableMemory shows available memory; a decreasing trend may suggest a leak. Option B is incorrect because CPUUtilization is not a direct memory metric. Option C is incorrect because ReadIOPS relates to I/O operations, not memory.

211
MCQhard

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?

A.Create a read replica and redirect all SELECT queries to the replica.
B.Enable Multi-AZ to offload CPU to the standby instance.
C.Increase the allocated storage to 1 TB to improve I/O performance.
D.Upgrade to a larger instance type, such as db.r5.xlarge.
AnswerD

A larger instance provides more CPU cores and better performance, directly addressing high CPU utilization.

Why this answer

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.

Exam trap

AWS often tests the misconception that increasing storage or adding a read replica can solve CPU bottlenecks, but the correct answer requires recognizing that CPU saturation is a compute issue best addressed by scaling instance size.

How to eliminate wrong answers

Option A is wrong because creating a read replica and redirecting SELECT queries offloads read traffic from the primary instance, but the primary still handles all DML and writes; if the CPU bottleneck is from both SELECT and DML processing on the primary, the replica does not reduce the primary's CPU load. Option B is wrong because Multi-AZ provides a standby instance for failover only; the standby does not serve read traffic or offload CPU from the primary—it is a synchronous replica that is not active for queries. Option C is wrong because increasing gp2 storage to 1 TB increases baseline IOPS from 1500 to 3000, which improves I/O throughput, but the problem is CPU utilization, not I/O latency or throughput; the instance is compute-bound, not storage-bound.

212
MCQhard

A company is running an Amazon RDS for SQL Server instance with Multi-AZ deployment. The database is used by a critical application. During a recent failover test, the application experienced a 2-minute downtime. The application's connection string uses the DB instance endpoint, not the cluster endpoint. Which change would minimize downtime during future failovers?

A.Modify the application to use the cluster endpoint instead of the instance endpoint
B.Increase the DB instance class size
C.Create a read replica in a different Availability Zone
D.Enable Multi-AZ on the DB instance
AnswerA

The cluster endpoint points to the current primary and updates automatically after failover, reducing downtime.

Why this answer

Using the cluster endpoint (the DNS name that automatically points to the current primary after failover) allows the application to reconnect to the new primary without waiting for DNS propagation, thus minimizing downtime. Option B is incorrect because increasing the DB instance class size does not reduce failover time. Option C is incorrect because read replicas are used for read scaling, not for reducing failover downtime.

Option D is incorrect because Multi-AZ is already enabled, so enabling it again would have no effect.

213
Multi-Selectmedium

A company is using Amazon DynamoDB and wants to monitor the read/write capacity utilization of a table. Which ONE AWS service can be used to set up alarms for capacity consumption?

Select 1 answer
A.Amazon DynamoDB Auto Scaling
B.AWS CloudTrail
C.Amazon CloudWatch Logs
D.Amazon CloudWatch
E.AWS Config
AnswersD

Amazon CloudWatch provides metrics for DynamoDB read/write capacity consumption and allows you to set alarms on these metrics. It is the direct and correct service for monitoring capacity utilization.

Why this answer

Amazon CloudWatch (D) is the primary service for monitoring DynamoDB capacity metrics such as ConsumedReadCapacityUnits and ConsumedWriteCapacityUnits, and it can trigger alarms based on thresholds. Option C (Amazon CloudWatch Logs) is not used for direct capacity monitoring because DynamoDB does not emit capacity consumption logs; those metrics are available only through CloudWatch Metrics. DynamoDB Auto Scaling (A) adjusts capacity but does not monitor or alarm.

AWS CloudTrail (B) records API calls, not capacity metrics. AWS Config (E) tracks configuration changes, not utilization.

214
Multi-Selectmedium

A company is using Amazon Redshift and notices that queries are running slowly. Which TWO system views should be used to identify the cause of the slow queries? (Choose TWO.)

Select 2 answers
A.STV_TBL_PERM
B.PG_TABLE_DEF
C.STL_QUERY
D.STL_LOAD_COMMITS
E.SVV_QUERY_STATE
AnswersC, E

STL_QUERY logs all queries with execution details, useful for identifying slow queries after they have run.

Why this answer

The correct answers are C and E. STL_QUERY contains detailed logs of all queries, including execution times, enabling identification of slow queries after execution. SVV_QUERY_STATE shows currently running queries, helping identify those that are running slowly.

Option A (STV_TBL_PERM) is for table permissions, not performance. Option B (PG_TABLE_DEF) shows table definitions. Option D (STL_LOAD_COMMITS) shows load commit information.

215
MCQeasy

A developer is receiving timeout errors when connecting to an Amazon ElastiCache for Redis cluster from an Amazon EC2 instance. The security group for the EC2 instance allows outbound traffic to the Redis cluster's security group on port 6379. The Redis cluster's security group does not allow inbound traffic from the EC2 instance. What is the most likely cause of the timeout?

A.The network ACL for the Redis subnet is blocking inbound traffic on port 6379
B.The Redis cluster security group does not have an inbound rule allowing traffic from the EC2 security group on port 6379
C.The subnet route table does not have a route to the Redis cluster
D.The Redis cluster is not accessible from within the same VPC
AnswerB

Inbound rules are required for the target security group.

Why this answer

Security groups are stateful, meaning that if an outbound rule allows traffic, the return traffic is automatically allowed. However, inbound traffic must be explicitly permitted by the target security group. In this scenario, the EC2 instance's security group allows outbound traffic to the Redis cluster's security group on port 6379, but the Redis cluster's security group does not have an inbound rule allowing traffic from the EC2 security group on that port.

Therefore, the connection is blocked, causing timeout errors. Option A is incorrect because network ACLs are stateless and would affect both inbound and outbound traffic, but the issue is specifically with security groups. Option C is incorrect because subnet route tables control network routing, not port-level access.

Option D is incorrect because ElastiCache for Redis clusters are accessible within the same VPC if proper security group rules are configured.

216
Multi-Selecthard

A company is using Amazon DynamoDB with a global table in two regions. The application is experiencing high write latency on the replica table in the secondary region. Which THREE factors could contribute to this issue?

Select 3 answers
A.Large item sizes being written to the table.
B.Network latency between the primary and secondary regions.
C.Auto scaling configuration on the replica table.
D.Low read capacity on the replica table.
E.Insufficient write capacity on the replica table.
AnswersA, B, E

Larger items take longer to replicate.

Why this answer

Global tables replicate writes asynchronously. Large item sizes (A) increase the time to process each write. Network latency (B) between regions affects the replication time.

Insufficient write capacity (E) on the replica table can cause throttling and increased latency. Auto scaling configuration (C) is not a direct cause; it helps manage capacity. Read capacity (D) does not affect write latency.

217
MCQhard

A team is troubleshooting a DynamoDB table that has high read latency. The table uses on-demand capacity and has a global secondary index (GSI). Which configuration is MOST likely causing the issue?

A.The GSI has provisioned capacity set too low
B.Time-to-Live (TTL) is enabled
C.DAX is enabled for the table
D.The table uses on-demand capacity
AnswerA

GSIs have independent capacity; throttling on GSI causes high latency.

Why this answer

An under-provisioned GSI can throttle reads even if the base table uses on-demand. Option B is wrong because TTL does not affect read latency. Option C is wrong because DAX reduces latency, not increases.

Option D is wrong because on-demand capacity handles bursts and does not cause high latency.

218
MCQhard

A company has an Amazon Aurora MySQL DB cluster with a primary instance and two Aurora Replicas. The application is experiencing high write latency. The primary instance's CPU utilization is at 90%, while the replicas are at 30%. The DB cluster parameter group has the default values. Which change is most likely to reduce write latency?

A.Increase the DB instance class of the primary instance.
B.Add more Aurora Replicas to distribute the write load.
C.Disable the binary log (binlog) on the DB cluster.
D.Increase the allocated storage of the cluster.
AnswerC

Binary logging adds CPU overhead; disabling it can reduce write latency.

Why this answer

High CPU utilization on the primary instance is often caused by the binary log (binlog) being enabled, which is enabled by default for Aurora MySQL. Binlog generation adds overhead to write operations. Disabling binlog reduces CPU usage on the primary, thereby reducing write latency.

Increasing the instance class (Option A) could help but is not as directly targeted as disabling binlog. Adding more Aurora Replicas (Option B) does not reduce write latency on the primary because replicas handle read traffic only. Increasing allocated storage (Option D) does not affect CPU or write latency.

219
MCQmedium

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?

A.Enable DynamoDB Accelerator (DAX) for the table.
B.Create a global secondary index on the table.
C.Increase the provisioned write capacity for the table.
D.Switch from eventually consistent reads to strongly consistent reads.
AnswerC

Increasing write capacity units reduces throttling for write requests.

Why this answer

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.

Exam trap

The trap here is that candidates may confuse read performance solutions (DAX, consistency changes) with write throttling issues, or incorrectly assume that adding a GSI will offload write traffic, when in fact it increases the write capacity burden on the base table.

How to eliminate wrong answers

Option A is wrong because DynamoDB Accelerator (DAX) is an in-memory cache that reduces read latency, not write throttling; it does not increase write capacity or reduce write request throttling. Option B is wrong because creating a global secondary index (GSI) does not reduce write throttling on the base table; in fact, GSIs consume additional write capacity from the base table's provisioned throughput, potentially worsening throttling. Option D is wrong because switching from eventually consistent reads to strongly consistent reads doubles the read capacity consumption per read request, increasing read throttling risk and having no effect on write throttling.

220
MCQmedium

An application running on Amazon EC2 is unable to connect to an Amazon RDS for SQL Server DB instance. The security group for the RDS instance allows inbound traffic from the security group of the EC2 instance on port 1433. The network ACLs allow all traffic. What is a likely cause of the connectivity issue?

A.The RDS instance is in a private subnet and does not have a public IP address
B.The network ACL is blocking the traffic
C.The security group inbound rule is incorrectly configured
D.The database port is not 1433
AnswerA

Without a public IP, the EC2 instance cannot reach it over the internet.

Why this answer

If the RDS instance is in a private subnet without a public IP address, the EC2 instance cannot connect using the public DNS endpoint unless the EC2 is in the same VPC. Since the security group and network ACLs are correctly configured, the most likely cause is that the RDS instance is not publicly accessible and the EC2 instance is attempting to connect from outside the VPC or via the public endpoint. Option B is wrong because network ACLs allow all traffic.

Option C is wrong because the security group inbound rule is correctly configured to allow traffic from the EC2 security group on port 1433. Option D is wrong because port 1433 is the default SQL Server port, and if it were not, the error message would indicate a timeout, not a connectivity issue.

221
Multi-Selecthard

Which THREE actions should be taken to troubleshoot a high number of ThrottlingExceptions from Amazon DynamoDB? (Choose 3.)

Select 3 answers
A.Examine the ConsumedWriteCapacity and ThrottledWriteCount metrics in CloudWatch
B.Enable DynamoDB Streams to offload writes
C.Implement exponential backoff in the application
D.Increase the write capacity units for the table
E.Change the read consistency to eventual
AnswersA, C, D

Helps identify if capacity is exceeded.

Why this answer

A, C, D are correct. Examining CloudWatch metrics helps identify throttling patterns. Implementing exponential backoff is a best practice.

Increasing provisioned capacity resolves throttling. B is wrong because enabling DynamoDB Streams does not affect throttling. E is wrong because changing consistency model does not affect write throttling.

222
MCQeasy

A database administrator is troubleshooting an issue where an Amazon RDS for PostgreSQL DB instance is not allowing connections. The administrator checks the security group and network ACLs, and they are correctly configured. What is the next step to diagnose the issue?

A.Reboot the DB instance
B.Modify the DB instance's parameter group
C.Create a DB snapshot
D.Review the DB instance error logs in Amazon CloudWatch
AnswerD

Logs can show reason for connection failures.

Why this answer

Reviewing the DB instance error logs in Amazon CloudWatch can reveal connection issues such as authentication failures, SSL/TLS errors, or the maximum number of connections being reached. This is a direct diagnostic step. Option A is incorrect because rebooting may temporarily resolve a connection issue but does not help identify the root cause.

Option B is incorrect because modifying the parameter group changes configuration settings, which is not a diagnostic action for existing connection problems. Option C is incorrect because creating a snapshot is for backup and recovery, not for troubleshooting current connectivity.

223
MCQhard

A company is using Amazon Redshift for data warehousing. The VACUUM operation is taking longer than expected, and the database administrator wants to identify the tables that require the most vacuuming effort. Which system table should be queried to find the percentage of deleted rows per table?

A.STL_QUERY
B.STV_TBL_PERM
C.PG_TABLE_DEF
D.SVV_TABLE_INFO
AnswerD

SVV_TABLE_INFO includes columns for unsorted rows and tombstone blocks.

Why this answer

(SVV_TABLE_INFO) is correct because this system table contains tombstone and unsorted statistics, including the percentage of deleted rows per table, which helps identify tables requiring vacuuming. Option A (STL_QUERY) is incorrect because it stores query execution logs, not table statistics. Option B (STV_TBL_PERM) is incorrect as it provides information about permanent table block allocations but not deleted row ratios.

Option C (PG_TABLE_DEF) is incorrect because it holds table schema definitions only, not operational metrics.

224
Multi-Selectmedium

A company is using Amazon Redshift for analytics. The database administrator notices that some queries are slow and the system is running out of memory. Which THREE steps should the administrator take to improve performance?

Select 3 answers
A.Increase the node size (scale up) to get more memory per node
B.Optimize the table design by choosing appropriate distkeys and sortkeys
C.Add more nodes to the cluster to increase total memory
D.Run the VACUUM command to reclaim space from deleted rows
E.Configure workload management (WLM) to limit the number of concurrent queries
AnswersB, C, E

Better data distribution reduces memory usage during joins.

Why this answer

Scaling up increases memory per node but may not be sufficient; scaling out (adding nodes) is often more effective. Option B is correct because proper distribution and sort keys minimize data shuffling and improve query performance. Option C is correct because adding more nodes increases the cluster's total memory and compute capacity.

Option D is incorrect because VACUUM reclaims disk space, not memory. Option E is correct because configuring WLM limits concurrent queries, reducing memory contention.

225
Multi-Selecthard

A company is monitoring an Amazon RDS for Oracle instance. CloudWatch alarms show that FreeableMemory is consistently below 256 MB. The database has high read and write I/O. Which THREE steps should the database specialist take to diagnose the issue?

Select 3 answers
A.Check the MemoryPressure and LogFileSyncDuration metrics in CloudWatch.
B.Review the Oracle memory advisor (V$MEMORY_TARGET_ADVICE).
C.Enable storage auto scaling to increase allocated storage.
D.Increase the DB instance class to allocate more memory.
E.Query V$SGASTAT and V$PGASTAT to understand memory allocation.
AnswersA, B, E

These metrics indicate memory pressure and potential performance impact.

Why this answer

Options A, B, and E are correct diagnostic steps. A: Checking MemoryPressure and LogFileSyncDuration helps identify memory pressure and I/O bottlenecks. B: Reviewing V$MEMORY_TARGET_ADVICE provides Oracle's memory sizing recommendations.

E: Querying V$SGASTAT and V$PGASTAT reveals detailed memory allocation within SGA and PGA. Option C (enabling storage auto scaling) addresses storage, not memory. Option D (increasing instance class) is a remediation step, not a diagnostic step.

← PreviousPage 3 of 4 · 289 questions totalNext →

Ready to test yourself?

Try a timed practice session using only Troubleshooting questions.