Examkin

Soru

Zorluk: ZorOptimizing Compute and Storage Performance

A digital sports media company is preparing for a live-streamed championship event that is expected to attract millions of concurrent viewers within a 55-minute window. The application runs on Amazon EC2 instances in an Auto Scaling group (ASG) behind an Application Load Balancer (ALB). The application reads and writes user profile and viewing state data to an Amazon Aurora PostgreSQL DB cluster consisting of one writer instance in Availability Zone (AZ) A and one reader instance in AZ B.

During load testing, the following performance issues are observed:
- A sudden, massive flood of requests at the start of the event causes the ALB to drop connections and return HTTP 503 (Service Unavailable) errors before the ASG or ALB can scale out.
- The reporting and leaderboard dashboard queries, which are read-heavy, cause significant CPU utilization spikes on the primary writer instance because the application is currently configured to connect to the DB cluster endpoint for all operations.

Which combination of architectural modifications will optimize the compute and storage layers to handle the event's performance requirements?

  1. Submit a support ticket to AWS to pre-warm the Application Load Balancer based on the expected traffic profile. Update the dashboard application configuration to use the reader endpoint of the Aurora DB cluster to distribute the read queries across the available Aurora Replicas.Cevap
  2. B
    Configure a target tracking scaling policy for the Application Load Balancer based on the ALBRequestCountPerTarget metric. Update the dashboard application configuration to use the reader endpoint of the Aurora DB cluster to distribute the read queries across the available Aurora Replicas.
  3. C
    Submit a support ticket to AWS to pre-warm the Application Load Balancer based on the expected traffic profile. Modify the Aurora DB cluster to enable Multi-AZ deployment with a standby instance, and configure the dashboard application to use the standby instance IP address to serve read queries.
  4. D
    Configure a target tracking scaling policy for the Application Load Balancer based on the ALBRequestCountPerTarget metric. Modify the Aurora DB cluster to enable Multi-AZ deployment with a standby instance, and configure the dashboard application to use the standby instance IP address to serve read queries.

Cevap

Submit a support ticket to AWS to pre-warm the Application Load Balancer based on the expected traffic profile. Update the dashboard application configuration to use the reader endpoint of the Aurora DB cluster to distribute the read queries across the available Aurora Replicas.
The correct option addresses both the network/compute bottleneck at the ingress layer and the database compute bottleneck. Requesting AWS to pre-warm the ALB ensures the load balancer can handle millions of concurrent connections immediately. Utilizing the Aurora reader endpoint ensures that read queries are routed to the Aurora reader replica rather than the writer, relieving CPU strain on the database cluster's primary node.

Adım Adım Çözüm

1
Analyze the load profile and ALB scaling capabilities.
Identify that the expected traffic spike is instantaneous (55-minute window), meaning standard ALB scaling will be too slow, resulting in dropped requests.
ALBs scale in response to traffic over time; sudden, massive surges require pre-warming by AWS Support to pre-provision capacity.
2
Address the write instance CPU utilization bottleneck.
Observe that write and read workloads are sharing the same primary instance, leading to resource contention.
Separating read queries from write queries is a standard pattern to optimize database compute resource usage.
3
Redirect read queries using the Aurora reader endpoint.
Configure the read-heavy reporting application to connect to the cluster reader endpoint instead of the primary writer cluster endpoint.
The reader endpoint automatically load-balances connection requests across all available Aurora Replicas, offloading read operations from the writer instance.

Anahtar Kavram

Optimizing compute and storage performance by combining ELB pre-warming for sudden peak workloads and Aurora reader endpoints for database read scaling.
Bu soruyu puanla

Konular

Optimizing Compute and Storage Performance