Examkin

Soru

Zorluk: ZorPerformance and Scalability Optimization

A digital health platform hosts a virtual daily fitness challenge. Every day at exactly 12:00 PM UTC, the platform experiences a sudden surge of 1,000,0001,000,000 concurrent users synchronizing their wearable telemetry data and requesting a live leaderboard. This activity window lasts for 55 minutes, resulting in an immediate 1010-fold increase in HTTP POST and GET requests. The architecture consists of an Application Load Balancer (ALB) routing requests to an Amazon Elastic Container Service (ECS) cluster on AWS Fargate, backed by an Amazon Aurora PostgreSQL database. Which design strategy provides the most performant, scalable, and operationally efficient solution to handle this daily traffic spike without dropping requests?

  1. Configure scheduled scaling for the ECS tasks to scale out before 12:00 PM UTC. Submit a request to AWS Support to pre-warm the ALB prior to the daily challenge window. Configure Aurora Auto Scaling to add Aurora Replicas to the DB cluster, and direct leaderboard read queries to the reader endpoint.Cevap
  2. B
    Configure target tracking scaling policies for the ECS tasks based on CPU utilization. Rely on the ALB's native automatic scaling to handle the sudden surge in HTTP requests. Configure Aurora Auto Scaling to provision additional Aurora Replicas, and direct leaderboard read queries to the reader endpoint.
  3. C
    Configure scheduled scaling for the ECS tasks to scale out before 12:00 PM UTC. Submit a request to AWS Support to pre-warm the ALB prior to the daily challenge window. Enable Aurora Multi-AZ deployment and configure the application to direct all leaderboard read queries to the standby DB instance to handle the read workload.
  4. D
    Configure scheduled scaling for the ECS tasks to scale out before 12:00 PM UTC. Submit a request to AWS Support to pre-warm the ALB prior to the daily challenge window. Deploy an Amazon ElastiCache for Memcached cluster with multi-AZ replication enabled to cache the leaderboard queries and ensure session persistence during the spike.

Cevap

Configure scheduled scaling for the ECS tasks to scale out before the event, request AWS Support to pre-warm the ALB, and configure Aurora Auto Scaling with Aurora Replicas while directing read queries to the reader endpoint.
The correct strategy uses scheduled scaling for ECS tasks and pre-warming for the ALB to handle the sudden, predictable spike, while scaling Aurora reads using Aurora Replicas via the reader endpoint.

Adım Adım Çözüm

1
Address the sudden network ingress spike at the load balancer tier.
Contact AWS Support to pre-warm the ALB so it is provisioned with sufficient capacity before the surge.
Standard ALB auto-scaling is reactive and cannot scale fast enough to prevent connection timeouts during an immediate 10-fold surge.
2
Address the compute scaling latency for container tasks.
Apply scheduled scaling to the ECS service to pre-provision Fargate tasks before 12:00 PM UTC.
Target tracking scaling policies take several minutes to detect high CPU and launch new tasks, which would fail to handle the initial minutes of the 5-minute peak.
3
Address the database read scaling requirements.
Configure Aurora Auto Scaling with Aurora Replicas and route read queries to the Aurora Reader endpoint.
Aurora Replicas handle read queries horizontally, unlike Multi-AZ standby instances which do not accept read traffic.

Anahtar Kavram

Handling predictable flash traffic spikes requires proactive resource provisioning at the load balancer and compute tiers, combined with horizontal read replica scaling at the database tier.
Tahmini Süre:2m 30s
Bu soruyu puanla

Konular

Performance and Scalability Optimization

Bu konudan daha fazla soru