A media company hosts a video processing platform on AWS. The application runs on Amazon E… — AWS Certified Solutions Architect - Professional

A media company hosts a video processing platform on AWS. The application runs on Amazon EC2 instances in an Auto Scaling Group (ASG) behind an Application Load Balancer (ALB). The instances process large 4K video uploads (averaging $40\text{ GB}$ to $60\text{ GB}$ each) that are downloaded from Amazon S3, transcoded locally, and uploaded back to S3. The transcoding software requires metadata from an Amazon Aurora PostgreSQL database deployed in a Multi-AZ configuration.

During recent scheduled marketing campaigns, the company encountered several performance issues:
* Transcoding execution times increased significantly, and the EC2 instances showed high I/O wait times and disk queue length, while instance CPU utilization remained under $35\%$ . The instances utilize $1\text{ TB}$ Amazon EBS gp3 volumes with default settings ( $3,000$ IOPS and $125\text{ MB/s}$ throughput).
* The ALB dropped incoming user connections at the start of the campaigns before the ASG could launch new instances.
* The database experienced high CPU utilization and query timeouts due to a spike in read-intensive analytical queries.

Which combination of actions will optimize the performance of the system to resolve these bottlenecks?

Increase the throughput of the gp3 volumes to $500\text{ MB/s}$ to resolve the storage bottleneck. Submit a request to AWS Support to pre-warm the Application Load Balancer prior to the scheduled marketing campaigns. Create Aurora Replicas to offload the read-intensive queries and update the application to use the Aurora reader endpoint.Cevap
B
Increase the throughput of the gp3 volumes to $500\text{ MB/s}$ to resolve the storage bottleneck. Submit a request to AWS Support to pre-warm the Application Load Balancer prior to the scheduled marketing campaigns. Configure the application to route read-intensive queries to the passive Multi-AZ standby instance to offload the primary instance.
C
Increase the throughput of the gp3 volumes to $500\text{ MB/s}$ to resolve the storage bottleneck. Configure the Auto Scaling Group to use step scaling policies with a lower CPU utilization threshold to quickly scale out the EC2 instances at the start of the campaigns. Create Aurora Replicas to offload the read-intensive queries and update the application to use the Aurora reader endpoint.
D
Replace the gp3 volumes with io2 volumes configured with $3,000$ Provisioned IOPS. Configure the Auto Scaling Group to use step scaling policies with a lower CPU utilization threshold to quickly scale out the EC2 instances. Configure the application to route read-intensive queries to the passive Multi-AZ standby instance.

Cevap

Increase the throughput of the gp3 volumes to

500\text{ MB/s}

, request ALB pre-warming from AWS Support prior to the campaigns, and add Aurora Replicas to serve read-intensive queries via the reader endpoint.

The correct option addresses the throughput bottleneck on the EBS volumes by increasing the gp3 throughput to

500\text{ MB/s}

(since large video file processing is highly sequential and throughput-bound rather than IOPS-bound). It solves the ALB flash traffic connection drops by proactively pre-warming the ALB via AWS Support. Lastly, it resolves the database query timeouts by scaling the database reads horizontally using Aurora Replicas and routing them to the reader endpoint.

Adım Adım Çözüm

Identify the EBS storage bottleneck.

The

1\text{ TB}

gp3 volume provides

125\text{ MB/s}

throughput by default. Transcoding large files (up to

60\text{ GB}

) is highly sequential and throughput-bound, causing high disk queue length and I/O wait times despite low CPU.

Increasing gp3 throughput to

500\text{ MB/s}

removes this bottleneck without changing the volume type.

Address the load balancer capacity during flash traffic.

ALB auto-scaling is gradual and cannot handle sudden 10x spikes instantly, causing connection drops.

Pre-warming the ALB via AWS Support scales the load balancer capacity in advance of the campaign.

Scale the database read throughput.

The database runs analytical queries on a Multi-AZ deployment.

Multi-AZ standby instances are passive and cannot serve read traffic. Adding Aurora Replicas and using the reader endpoint scales reads horizontally.

Anahtar Kavram

Identifying and optimizing compute, storage, and load balancing bottlenecks by utilizing GP3 throughput scaling, ALB pre-warming for flash traffic, and Aurora Replicas for read scalability.

Soru

Cevap

Adım Adım Çözüm

Anahtar Kavram

Sık Yapılan Hatalar

Konular