Examkin

Question

Difficulty: Very hardPerformance and Scalability Optimization

A global virtual event management company is deploying a live interactive polling application. The application will host a synchronous global event at exactly 12:00 UTC, where a baseline of 5,0005,000 concurrent HTTP connections is expected to immediately spike to 2,000,0002,000,000 concurrent connections within a 6060-second window. The application is hosted on Amazon ECS using the AWS Fargate launch type, running behind an Application Load Balancer (ALB). The containerized application reads event metadata from an Amazon Aurora PostgreSQL database cluster. The database must maintain sub-10-millisecond response times under peak load. Which combination of scaling and configuration actions should a solutions architect recommend to ensure the application scales seamlessly during the launch while maintaining performance?

  1. Request ALB pre-warming from AWS Support based on the expected traffic profile. Configure ECS Scheduled Scaling to provision the target number of Fargate tasks 30 minutes prior to the event. Implement Aurora Auto Scaling for the database cluster's Aurora Replicas to scale out read capacity in response to the query load.Answer
  2. B
    Rely on the ALB's native automatic scaling to handle the connection spike dynamically. Configure ECS Service Auto Scaling with target tracking policies based on average CPU utilization. Configure Aurora Auto Scaling for the database cluster's Aurora Replicas.
  3. C
    Request ALB pre-warming from AWS Support based on the expected traffic profile. Configure ECS Scheduled Scaling to provision the target number of Fargate tasks 30 minutes prior to the event. Rely on the Amazon RDS PostgreSQL Multi-AZ standby instance to automatically scale and absorb the incoming read query load.
  4. D
    Configure an Amazon CloudFront distribution in front of the ALB with custom caching behaviors. Configure ECS Service Auto Scaling to scale based on CPU utilization. Implement Amazon RDS Multi-AZ with two readable standbys and configure the application to load balance database queries across them.

Answer

Request ALB pre-warming from AWS Support based on the expected traffic profile. Configure ECS Scheduled Scaling to provision the target number of Fargate tasks 30 minutes prior to the event. Implement Aurora Auto Scaling for the database cluster's Aurora Replicas to scale out read capacity in response to the query load.
The correct option addresses the scaling challenges at every layer of the architecture for a highly predictable flash-traffic event. Requesting ALB pre-warming ensures that the load balancer is scaled in advance to handle the 2,000,0002,000,000 concurrent connections. Configuring ECS Scheduled Scaling launches Fargate tasks 30 minutes early, removing the container bootstrapping delay. Finally, utilizing Aurora Auto Scaling dynamically manages database read replicas to handle query loads, ensuring sub-10ms response times.

Step-by-Step Solution

1
Analyze the load balancer scaling characteristics.
Determine that an Application Load Balancer scales in response to traffic but requires time to provision resources. A sudden, multi-magnitude spike over a 60-second window will exceed the ALB's dynamic scaling rate, requiring manual pre-warming.
AWS Support must pre-warm the ALB based on the expected Request Per Second (RPS) and connection profile to prevent dropped packets.
2
Evaluate application-tier container provisioning latency.
Identify that dynamic Auto Scaling policies (like target tracking) take time to trigger, pull images, and bootstrap container tasks.
ECS Scheduled Scaling is required to pre-provision the ECS Fargate tasks before the peak event starts at 12:00 UTC.
3
Evaluate database horizontal scalability.
Confirm that Aurora PostgreSQL read traffic must scale horizontally using Aurora Replicas under an Auto Scaling policy.
Standard Multi-AZ standby instances do not serve read requests and cannot scale read traffic. Aurora Replicas must be scaled to maintain sub-10ms query times.

Key Concept

Handling flash traffic spikes by combining load balancer pre-warming, scheduled compute capacity scaling, and horizontal database read replicas.
Estimated Time:3m 0s
Rate this question

Topics

Performance and Scalability Optimization

More questions from this topic