Examkin

Soru

Zorluk: Çok zorPerformance and Scalability Optimization

An advertising technology platform is preparing to launch a real-time bidding service that receives bid requests from thousands of publishers. The service is expected to experience a baseline traffic of 20,00020,000 requests per second (RPS), with immediate spikes up to 300,000300,000 RPS during scheduled regional shopping events. The backend runs on Amazon Elastic Container Service (ECS) on AWS Fargate behind an Application Load Balancer (ALB). The application queries an Amazon Aurora PostgreSQL database for campaigns and user profiles. To ensure sub-second response times and high availability during these flash traffic events, which two architectural optimization strategies should a Solutions Architect recommend?

  1. Submit a support case to AWS to pre-warm the Application Load Balancer (ALB) with the expected traffic profile and rate of increase before the scheduled shopping events.Cevap
  2. Configure Application Auto Scaling for Aurora Replicas to dynamically scale read capacity based on average CPU utilization, and configure the application to use the Aurora reader endpoint for read queries.Cevap
  3. C
    Rely on the Application Load Balancer's (ALB) built-in automatic scaling to handle the sudden 300,000300,000 RPS traffic spikes without pre-warming, as ALBs dynamically scale to any load instantly.
  4. D
    Configure the Amazon Aurora database with a Multi-AZ deployment and configure the application to route read queries to the passive standby instance to offload the primary instance.
  5. E
    Set the scale-out cooldown period for the ECS service's Auto Scaling policy to 1010 seconds to ensure tasks scale out rapidly without waiting for previous tasks to finish launching.

Cevap

To handle sudden, massive traffic spikes and maintain database performance, the solutions architect should request ALB pre-warming from AWS Support to prevent initial connection drops, and configure dynamic Application Auto Scaling for Aurora Replicas to scale read queries distributed through the Aurora reader endpoint.
Requesting ALB pre-warming prepares the load balancer infrastructure to accept 300,000300,000 RPS without dropping connections due to rapid traffic surges. Using Aurora Replicas with Auto Scaling and targeting the reader endpoint dynamically spreads the query load across active readers, preventing database bottlenecking.

Adım Adım Çözüm

1
Evaluate the ingress layer scaling characteristics during instantaneous flash traffic.
The traffic spikes instantly from 20,00020,000 to 300,000300,000 RPS, exceeding the standard ALB scaling rate limit. AWS Support must pre-warm the ALB.
Prevents dropped connections and 502502/503503 errors during the initial phase of the flash traffic event.
2
Address read scalability constraints at the database tier.
Determine that Aurora Replicas can scale horizontally and handle read traffic using the cluster's reader endpoint.
Offloads the primary database writer node and scales query capacity dynamically using CPU utilization as the target metric.
3
Analyze ECS Fargate auto-scaling policies for transient periods.
Ensure the scale-out cooldown period is configured to accommodate Fargate task provisioning and bootstrapping times.
Prevents rapid, redundant scale-out alarms (thrashing) that happen when the cooldown expires before the new task is fully active and taking load.

Anahtar Kavram

Handling massive flash traffic spikes requires pre-warming load balancing resources, scaling read replicas dynamically using correct database endpoints, and configuring container scaling metrics with realistic cooldown targets.
Bu soruyu puanla

Konular

Performance and Scalability Optimization

Bu konudan daha fazla soru