Examkin

Soru

Zorluk: ZorOptimizing Compute and Storage Performance

A major media company operates a digital ticket-booking application. The application's web tier runs on Amazon EC2 instances in an Auto Scaling group behind an Application Load Balancer (ALB). The database tier runs on an Amazon RDS for PostgreSQL DB instance configured in a Multi-AZ deployment. The company is preparing for a flash-sale event where traffic is expected to increase by 200×200 \times within 22 minutes. During a previous similar event, users experienced connection timeouts and HTTP 503 Service Unavailable errors. Monitoring showed that:

1. The ALB dropped requests during the first 55 minutes of the event.
2. The EC2 Auto Scaling group took over 1010 minutes to provision and bootstrap new instances due to a complex user data script that downloads dependencies and updates application assets from Amazon S3.
3. The database CPU utilization reached 100%100\% due to a high volume of read-only queries for event details.

Which set of actions should a Solutions Architect recommend to optimize the performance of the compute and storage tiers for the upcoming event?

  1. Submit a request to AWS Support to pre-warm the Application Load Balancer prior to the event. Configure the EC2 Auto Scaling group to use a custom AMI with all dependencies pre-installed and maintain a warm pool of pre-warmed instances. Deploy an Amazon RDS for PostgreSQL read replica to offload the read-only query traffic from the primary DB instance.Cevap
  2. B
    Submit a request to AWS Support to pre-warm the Application Load Balancer. Configure the EC2 Auto Scaling group to use a custom AMI with all dependencies pre-installed and maintain a warm pool of pre-warmed instances. Configure the application to route the read-only queries to the standby DB instance in the secondary Availability Zone of the RDS Multi-AZ deployment.
  3. C
    Rely on the Application Load Balancer's automatic scaling to dynamically scale out the web tier. Configure the EC2 Auto Scaling group to use a custom AMI with all dependencies pre-installed and maintain a warm pool of pre-warmed instances. Deploy an Amazon RDS for PostgreSQL read replica to offload the read-only query traffic from the primary DB instance.
  4. D
    Rely on the Application Load Balancer's automatic scaling to dynamically scale out the web tier. Configure the EC2 Auto Scaling group to use a launch template that executes a user data script on new instances. Route the read-only queries to the standby DB instance in the secondary Availability Zone of the RDS Multi-AZ deployment.

Cevap

Submit a request to AWS Support to pre-warm the Application Load Balancer prior to the event. Configure the EC2 Auto Scaling group to use a custom AMI with all dependencies pre-installed and maintain a warm pool of pre-warmed instances. Deploy an Amazon RDS for PostgreSQL read replica to offload the read-only query traffic from the primary DB instance.
The correct solution optimizes the entire architecture. For the web tier, pre-warming the ALB ensures it does not drop connections during the sudden, massive traffic spike, while utilizing custom AMIs and warm pools eliminates the 1010-minute bootstrapping delay of the instances. For the database tier, deploying an RDS read replica offloads the heavy read query volume from the primary DB instance, which resolves the database CPU exhaustion.

Adım Adım Çözüm

1
Address the entry-point traffic bottleneck by submitting an AWS Support request to pre-warm the Application Load Balancer.
The ALB has sufficient capacity provisioned upfront to handle the sudden 200×200 \times traffic increase without dropping connections.
Default ALB auto-scaling is reactive and cannot scale fast enough for near-instantaneous spikes.
2
Optimize the compute boot performance by creating a custom AMI containing all required application dependencies and utilizing EC2 Auto Scaling Warm Pools.
EC2 instances can join the active cluster immediately without waiting for user data execution, reducing scaling latency from over 1010 minutes to under a minute.
Downloading files and installing software on boot introduces severe bootstrapping delay under sudden load.
3
Offload read-only query traffic from the primary RDS PostgreSQL DB instance by deploying a read replica.
The read replica handles the read-only queries, reducing the CPU utilization on the primary database and ensuring write operations do not time out.
An RDS Multi-AZ standby DB instance cannot receive client query traffic; only dedicated read replicas scale read workloads.

Anahtar Kavram

Handling extreme flash traffic spikes requires a combination of proactive load balancer pre-warming, pre-baked compute resources (custom AMIs and warm pools) to avoid bootstrap latency, and read replicas to scale storage read operations.
Bu soruyu puanla

Konular

Optimizing Compute and Storage Performance