Examkin

Soru

Zorluk: OrtaOptimizing Compute and Storage Performance

A gaming company operates a multiplayer online game where player session state and matchmaking queues are managed by an application running on Amazon EC2 instances in an Auto Scaling group behind an Application Load Balancer (ALB). The backend database is an Amazon RDS for PostgreSQL DB instance configured in a Multi-AZ deployment. During daily peak hours, players experience severe latency spikes and disconnected sessions. Monitoring data shows that the RDS primary DB instance has 90% CPU utilization and read IOPS throttling while the standby DB instance has 0% CPU utilization. Additionally, the EC2 instances experience a sudden CPU utilization spike of 98% for the first 5 minutes of the peak window, causing the ALB to report 502 Bad Gateway errors before the Auto Scaling group can successfully launch and bootstrap new instances. Which combination of actions should a Solutions Architect recommend to resolve the database and compute performance bottlenecks?

  1. A
    Deploy Amazon RDS Read Replicas to scale read queries, and request AWS Support to pre-warm the Application Load Balancer to handle the initial traffic spikes.
  2. B
    Route player session read queries to the passive standby DB instance in the Multi-AZ deployment, and configure a warm pool for the Auto Scaling group.
  3. Deploy Amazon RDS Read Replicas to offload the read-heavy matchmaking queries from the primary DB instance, and configure the Auto Scaling group with a warm pool and step scaling policies.Cevap
  4. D
    Increase the Auto Scaling group default cooldown period to prevent rapid scaling actions, and upgrade the RDS instance to a larger instance class with Multi-AZ enabled.

Cevap

Deploy Amazon RDS Read Replicas to offload the read-heavy matchmaking queries from the primary DB instance, and configure the Auto Scaling group with a warm pool and step scaling policies.
Deploying Amazon RDS Read Replicas is the correct approach to scale read workloads because it allows the primary instance to delegate read queries to replicas. Utilizing an Auto Scaling group warm pool keeps instances in a stopped or running state with pre-bootstrapped applications, drastically reducing the time it takes to serve traffic during sudden scaling events.

Adım Adım Çözüm

1
Identify the database bottleneck and distinguish between Multi-AZ and Read Replica functionalities.
Recognize that the standby database in an RDS Multi-AZ configuration is passive and cannot serve read traffic.
This rules out solutions attempting to route reads to the standby DB instance.
2
Address the database read scaling requirements.
Deploy Amazon RDS Read Replicas to offload read-heavy query workloads from the primary DB instance.
Read replicas are designed specifically to scale read throughput horizontally.
3
Address the EC2 compute scaling bottleneck.
Implement an Auto Scaling group warm pool and step scaling policies to keep instances in a pre-bootstrapped state.
This avoids the 5-minute cold start/bootstrapping latency that causes the initial CPU spike and ALB 502 errors.

Anahtar Kavram

Scaling read workloads in Amazon RDS using Read Replicas vs Multi-AZ, and optimizing EC2 Auto Scaling performance using warm pools and scaling policies.
Tahmini Süre:2m 0s
Bu soruyu puanla

Konular

Optimizing Compute and Storage Performance