SOA-C02 Cost and Performance Optimization Practice Test 2 — 15 Questions

Question 1

A company uses Amazon S3 to store backups and logs. The total data volume is 50 TB and grows by 2 TB per month. The company requires that data be retained for 7 years. Which THREE actions will optimize storage costs? (Choose THREE.)

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

Accepted Answer

Use S3 Lifecycle policies to transition data to S3 Glacier Deep Archive after 30 days.. The correct answers are A, B, and D. S3 Lifecycle policies help optimize storage costs. Option A transitions data to S3 Glacier Deep Archive, the lowest-cost storage class for long-term retention. Option B expires incomplete multipart uploads, preventing unnecessary storage charges. Option D uses S3 Intelligent-Tiering, which automatically moves data between tiers based on changing access patterns, reducing costs for infrequently accessed data with unknown patterns. Option C is incorrect because deleting data after one year fails to meet the 7-year retention requirement. Option E is incorrect because S3 Glacier Flexible Retrieval is more expensive than S3 Glacier Deep Archive and is not necessary for data that is rarely accessed.

Answer

Store all data in S3 Standard for the first year, then delete.

Answer

Use S3 Glacier Flexible Retrieval to allow data retrieval in minutes.

Question 2

A SysOps administrator attached the IAM policy shown to a user. The user needs to launch EC2 instances for a cost-sensitive project. What is the impact of this policy on cost optimization?

Accepted Answer

The policy allows launching any number of t3.micro instances, potentially increasing costs if many instances are launched.. The policy allows launching any number of t3.micro instances because it does not include any conditions that limit the number or total cost. While t3.micro instances are relatively cheap, launching many of them can still increase costs significantly. Therefore, the policy does not enforce cost optimization beyond restricting instance type.

Answer

The policy is too restrictive because it only allows t3.micro instances, which may not be sufficient for production workloads.

Answer

The policy does not allow launching any instances because the condition is on RunInstances.

Answer

The policy ensures cost control by allowing only t3.micro instances, which are the cheapest.

Question 3

A SysOps administrator reviews the Lambda invocation metrics for a function. The function processes data from an S3 bucket and is triggered by S3 events. The cost for this function is higher than expected. What should the administrator do to reduce costs?

Accepted Answer

Increase the function's memory allocation to reduce execution time.. Lambda pricing is based on GB-seconds (memory × duration) plus invocation count. Increasing memory allocation gives the function more CPU proportionally, which can dramatically reduce execution time. If the function is CPU-bound, the total GB-seconds can decrease even though the per-millisecond rate is higher, lowering overall cost.

Answer

Set a reserved concurrency limit to cap the number of concurrent executions.

Answer

Change the function to be invoked on a schedule instead of by S3 events.

Answer

Reduce the function's memory allocation to lower cost per invocation.

Question 4

A company has a production RDS MySQL instance with 500 GB of gp2 storage. The database experiences high read traffic. The SysOps team needs to improve read performance at minimal cost. What should they do?

Accepted Answer

Create one or more RDS Read Replicas in the same region and direct read traffic to them.. Creating RDS Read Replicas in the same region offloads read queries from the primary instance, improving read performance cost-effectively. Horizontal scaling with replicas is typically cheaper and more scalable than vertical scaling. Option A is wrong: migrating to Aurora is more expensive and auto scaling is not the primary solution for read-heavy workloads. Option B is wrong: switching to gp3 and increasing IOPS improves storage performance, but the bottleneck is more likely database processing, not IOPS. Option C is wrong: upgrading to a larger instance type increases cost without the scalability of replicas and may require downtime.

Answer

Migrate the database to Amazon Aurora with Auto Scaling enabled.

Answer

Change the storage type to gp3 and increase IOPS.

Answer

Upgrade the DB instance to a larger instance type with more vCPUs.

Question 5

A SysOps administrator is investigating a cost increase in a production AWS account. They notice that an EC2 instance with a Reservation has been running continuously for months. The instance type is m5.large in us-east-1. The administrator sees that the instance is using a Standard Reserved Instance (RI) that was purchased 6 months ago for a 1-year term. However, the current utilization shows that the instance is only used for 4 hours per day. What should the administrator do to optimize costs without affecting availability?

Accepted Answer

Modify the existing RI to a smaller instance size (e.g., m5.large to m5.xlarge? no, smaller: e.g., t3.medium) and use Auto Scaling with a schedule to start/stop the instance during business hours.. By modifying the existing Standard RI to a smaller instance size (e.g., within the same family or by converting to a Convertible RI to allow family changes), the administrator can better match the actual low utilization pattern. Additionally, using Auto Scaling with a scheduled scaling policy to start and stop the instance during business hours ensures availability during needed times and reduces waste. Option A is incorrect because stopping the instance does not stop the RI charges; Reserved Instance benefits are applied only to running instances, so the RI would be wasted during off-hours. Option B is incorrect because selling the RI on the Reserved Instance Marketplace typically incurs a loss of the upfront payment and does not directly address the over-provisioning issue. Option D is incorrect because converting to a larger instance family would increase compute capacity per hour, leading to higher costs and even more waste given the low utilization.

Answer

Stop the instance during off-hours using Instance Scheduler, but keep the RI as-is.

Answer

Sell the current RI on the Reserved Instance Marketplace and purchase a new Convertible RI for a smaller instance type.

Answer

Convert the RI to a Convertible RI and exchange it for a larger instance family to get more compute per hour.

Question 6

A company uses Amazon CloudWatch Logs to store application logs. The logs are retained indefinitely. The compliance team requires logs to be retained for 7 years. What is the MOST cost-effective way to meet this requirement?

Accepted Answer

Set a log retention policy in CloudWatch Logs to expire logs after 7 years.. Setting a log retention policy in CloudWatch Logs is the most cost-effective way because it automatically deletes logs older than the specified period (7 years) without incurring additional storage or export costs. Option A is incorrect because CloudWatch Logs retention policy, not exporting to S3, is simpler and cheaper. Option B is incorrect because exporting to S3 incurs data transfer costs and S3 storage costs, and while Glacier Deep Archive is cheap, the lifecycle transition may incur costs and the total cost is higher than simple retention policy. Option C is incorrect because disabling logging after 7 years does not address the retention requirement and may violate compliance. Option D is correct as it directly meets the 7-year retention requirement with minimal cost and management overhead.

Answer

Export logs to Amazon S3 and immediately transition to S3 Glacier Deep Archive.

Answer

Export logs to Amazon S3 and create an S3 Lifecycle policy to transition objects to S3 Glacier Deep Archive after 7 years.

Answer

Disable logging after 7 years to stop incurring costs.

Question 7

A company uses this IAM policy to restrict S3 uploads to the STANDARD storage class only. A developer tries to upload an object to the bucket using the CLI without specifying a storage class. What will happen?

Accepted Answer

The upload will succeed because the default storage class is STANDARD.. The IAM policy condition requires s3:x-amz-storage-class to be STANDARD. When the developer does not specify a storage class in the CLI upload command, the default storage class STANDARD is applied, which matches the condition, so the upload succeeds. Option A is wrong because the condition does allow PutObject when the storage class is STANDARD. Option B is wrong because the bucket default storage class does not affect the IAM policy evaluation; the policy checks the actual storage class of the upload, which defaults to STANDARD. Option D is wrong because the condition does not require explicit specification; it only checks the value, and the default satisfies it.

Answer

The upload will fail because the condition does not allow PutObject.

Answer

The upload will succeed if the bucket has a default storage class of STANDARD-IA.

Answer

The upload will fail because the condition requires explicit specification of the storage class.

Question 8

A SysOps administrator is optimizing an AWS Lambda function that processes data from an S3 bucket. The function currently runs with 128 MB memory and a 5-second timeout. The function experiences timeouts for large files. The administrator wants to improve performance and reduce cost. Which TWO actions should they take?

Accepted Answer

Enable provisioned concurrency to reduce cold starts.. To improve performance and reduce cost for a Lambda function that times out on large files, the two best actions are enabling provisioned concurrency (B) to reduce cold starts, and increasing memory (E) to provide more CPU and potentially faster execution, which can reduce execution time and may lower cost due to faster processing. Option A is incorrect because increasing timeout to 15 minutes does not improve performance; it may increase cost if the function runs longer. Option C is incorrect because using S3 Standard-IA is for storage cost optimization, not for compute performance. Option D is incorrect because creating multiple Lambda functions does not address the performance of a single function and may increase complexity.

Answer

Increase the function timeout to 15 minutes.

Answer

Configure the S3 event notification to use S3 Standard-IA for the input files.

Answer

Create multiple Lambda functions to process different files concurrently.

Question 9

A company runs a web application on EC2 instances with an Auto Scaling group. The application uses an Amazon RDS MySQL database. During a traffic spike, the database CPU utilization reaches 90%. The SysOps team needs to improve database performance cost-effectively. Which THREE actions should they take?

Accepted Answer

Enable RDS Performance Insights to identify slow queries.. Option C is correct because RDS Performance Insights provides detailed database load metrics and helps the team pinpoint the specific SQL queries causing the 90% CPU utilization, enabling targeted optimization before spending on hardware. Option D is correct because adding RDS Read Replicas offloads read-only traffic from the primary MySQL instance, which is a cost-effective way to scale read-heavy web applications without over-provisioning the primary. Option E is correct because RDS Proxy pools and reuses database connections, reducing the connection overhead and CPU load on the RDS instance during traffic spikes, which improves performance without changing instance size. Option A is not correct because Multi-AZ is a high-availability and failover feature, not a performance-scaling mechanism, and it does not reduce CPU utilization on the primary. Option B is not correct because upgrading to a larger instance type may improve performance but is not the most cost-effective first action and does not address inefficient queries or connection overhead.

Answer

Enable Multi-AZ for the RDS instance.

Answer

Upgrade the RDS instance to a larger instance type.

Question 10

A company has three EC2 instances as shown in the exhibit. The company pays for a t3.micro Reserved Instance. Which instance(s) will receive the RI discount?

Accepted Answer

Only i-0abcd1234. Reserved Instance discounts apply only to instances that match the RI's attributes — instance type, platform, tenancy, and Availability Zone (for zonal RIs) or region (for regional RIs). A t3.micro RI applies to a single t3.micro instance in the specified scope. Only i-0abcd1234 matches the t3.micro instance type and the RI's scope, so only that instance receives the discount.

Answer

i-0abcd1234 and i-0abcd5678

Answer

i-0abcd1234 and i-0abcd9012

Answer

All three instances

Question 11

A SysOps administrator reviews the CloudWatch metrics for an EC2 instance as shown. The instance is part of an Auto Scaling group with a target tracking scaling policy based on average CPU utilization set to 50%. During the time period shown, what is the expected behavior?

Accepted Answer

The Auto Scaling group will scale out because CPU utilization exceeded 50% for several consecutive data points.. The target tracking scaling policy continuously adjusts capacity to keep the average CPU utilization at the target value of 50%. When the metric exceeds the target for a sustained period (as shown by several consecutive data points above 50%), the policy triggers a scale-out action to add instances and bring utilization back down. The fact that CPU later declines does not negate the earlier breach; the policy reacts to the breach as it occurs.

Answer

The Auto Scaling group will not scale because the average CPU utilization over the hour is below 50%.

Answer

The Auto Scaling group will scale in because the CPU utilization dropped below 50% at the end.

Answer

The Auto Scaling group will not scale because the metric shows a decline after the peak.

Question 12

A SysOps administrator receives an AWS Budgets alert that the monthly cost for an S3 bucket has increased by 300% compared to the previous month. The bucket stores log files and uses S3 Standard storage. The bucket has versioning enabled and lifecycle policies that transition objects to S3 Glacier Deep Archive after 90 days. What is the MOST likely cause of the cost increase?

Accepted Answer

The cost of lifecycle transition requests to Glacier Deep Archive.. The most likely cause of the cost increase is the cost of lifecycle transition requests to Glacier Deep Archive. When objects are transitioned, each transition request incurs a cost, and if there is a large number of objects or frequent transitions, this can significantly increase the bill. The question states that lifecycle policies transition objects after 90 days, so a sudden increase could be due to a batch of objects reaching that age and being transitioned.

Answer

The increased storage costs due to S3 Standard storage for the first 90 days.

Answer

Data transfer costs for downloading logs from the bucket.

Answer

The cost of storing multiple versions of objects due to versioning.

Question 13

A company runs a production database on a db.r5.large RDS instance with Multi-AZ enabled. They notice that the CPU utilization is consistently below 10% and memory usage is below 20%. The application is not expected to grow significantly. Which action would optimize costs without sacrificing high availability?

Accepted Answer

Change the instance type to a smaller size, such as db.t3.medium, and keep Multi-AZ.. The RDS instance is underutilized (CPU <10%, memory <20%) and the application is not expected to grow, so downsizing to a smaller instance type like db.t3.medium reduces cost while maintaining Multi-AZ for high availability. This directly addresses the cost optimization goal without sacrificing HA.

Answer

Remove Multi-AZ and use a single-AZ deployment to reduce costs.

Answer

Purchase a Reserved Instance for the current instance type to get a discount.

Answer

Change the instance type to a larger size, such as db.r5.xlarge, to improve performance.

Question 14

A company has a web application that uses an Application Load Balancer (ALB) and an Auto Scaling group of EC2 instances. The Auto Scaling group uses a dynamic scaling policy based on average CPU utilization. The SysOps administrator notices that during traffic spikes, the scaling policy adds instances too slowly, causing performance degradation. Which action should the administrator take to improve performance without significantly increasing costs?

Accepted Answer

Change the scaling policy to a target tracking policy with a target CPU utilization of 70%.. Target tracking scaling policies continuously adjust capacity to keep a metric at a target value, reacting faster than step or simple scaling because they use metric math and pre-defined thresholds internally. Setting a 70% CPU target gives headroom below the spike ceiling, so instances launch before CPU saturates. Unlike simple scaling, target tracking does not require a cooldown to trigger subsequent actions, so it responds to traffic spikes more quickly.

Answer

Use a scheduled scaling policy to add instances during known peak hours.

Answer

Increase the cooldown period for the scaling policy to allow instances to warm up.

Answer

Switch to a simple scaling policy that scales out when CPU exceeds 80%.

Question 15

A company is using Amazon CloudFront to deliver content to users worldwide. They notice high data transfer costs. Which THREE actions can reduce CloudFront data transfer costs?

Accepted Answer

Use a CloudFront price class that only uses the cheapest edge locations.. Option B is correct because selecting a CloudFront price class (e.g., Price Class 100, which uses only the least expensive North American and European edge locations) excludes costlier regions and directly lowers the per-GB data transfer rates charged for content delivery. Option C is correct because enabling automatic compression (gzip/Brotli) on compressible objects shrinks the payload size transferred from edge locations to viewers, reducing the number of bytes billed for data transfer out. Option E is correct because Origin Shield adds a centralized caching layer that consolidates origin fetches, raising the cache hit ratio and cutting the cache-miss traffic that CloudFront must pull from the origin, which reduces origin-side data transfer charges. Option A does not belong because Lambda@Edge runs custom compute at edge locations and adds request/execution charges rather than reducing data transfer costs. Option D does not belong because adding more edge locations does not lower transfer pricing and can even increase costs, since CloudFront already uses its full global network and the price class determines which locations are billed.

Answer

Use Lambda@Edge to customize content.

Answer

Add more edge locations to improve cache hit ratio.