Examkin

Soru

Zorluk: OrtaPerformance and Scalability Optimization

A global educational institution is launching a synchronized national online examination platform. At exactly 09:00 AM UTC, up to 250,000 students will simultaneously log in, retrieve their personalized exam papers, and start submitting their answers. The architecture will use an Application Load Balancer (ALB) to distribute traffic to Amazon ECS on AWS Fargate tasks, with Amazon Aurora PostgreSQL as the database. To handle this massive, instantaneous surge in traffic without dropping requests or causing database bottlenecks, which of the following actions should the Solutions Architect perform? (Select TWO.)

  1. Open a support ticket to pre-provision capacity for the Application Load Balancer (ALB) in anticipation of the 250,000 concurrent student connections.Cevap
  2. Establish a time-based scaling plan for the Amazon ECS service to provision the maximum required Fargate tasks 30 minutes prior to the exam start time.Cevap
  3. C
    Rely on the default self-scaling behavior of the Application Load Balancer (ALB) to dynamically adjust to the sudden influx of connections at the start of the exam.
  4. D
    Direct the exam paper retrieval read queries to the Aurora Multi-AZ standby instance to reduce the processing burden on the primary database instance.
  5. E
    Deploy an Amazon ElastiCache for Memcached cluster to act as a persistent database replication store for session states and ensure user sessions are not lost.

Cevap

The correct actions are to pre-provision Application Load Balancer (ALB) capacity by opening a support ticket and to establish a scheduled time-based scaling plan for the Amazon ECS service to spin up Fargate tasks before the exam begins.
To support a massive, synchronized event with 250,000 concurrent users at an exact start time, the architecture must pre-provision capacity. Pre-warming the Application Load Balancer ensures that the load balancing tier is ready to accept the sudden wave of connections without dropping requests. Using a scheduled time-based scaling policy for Amazon ECS Fargate tasks guarantees that the compute capacity is fully scaled out and operational before the load hits, preventing latency and request failures during task initialization.

Adım Adım Çözüm

1
Analyze the traffic pattern and workload characteristics.
The workload exhibits a massive, instant spike (flash traffic) of 250,000 users starting exactly at 09:00 AM UTC, which cannot be handled by dynamic reactive scaling.
Understanding the rapid ramp-up time is critical to choosing proactive scaling and pre-provisioning strategies over reactive ones.
2
Evaluate the scaling mechanics of the Application Load Balancer (ALB).
Standard ALBs scale progressively. For immediate spikes of this scale, requests will fail unless the ALB is pre-warmed.
Pre-warming ensures the load balancer's network capacity is configured for the expected surge in advance.
3
Evaluate the scaling mechanics of the compute tier (ECS on Fargate).
Dynamic scaling policies based on metrics like CPU utilization take several minutes to trigger and launch tasks, which is too slow. Scheduled scaling ensures the required Fargate tasks are running before the start time.
Proactive time-based scaling eliminates the startup delay associated with provisioning container capacity under load.

Anahtar Kavram

Handling instantaneous flash traffic spikes requires proactive capacity management, including pre-warming load balancers and using scheduled scaling policies for the compute tier, rather than relying on reactive scaling.
Bu soruyu puanla

Konular

Performance and Scalability Optimization

Bu konudan daha fazla soru