Examkin

Question

Difficulty: MediumOptimizing Compute and Storage Performance

An IoT company operates a real-time data ingestion pipeline on AWS. The pipeline uses an Auto Scaling group of Amazon EC2 instances to process incoming telemetry messages. The processed data is written to a shared Amazon EFS file system configured in Provisioned Throughput mode, while metadata is written to and queried from an Amazon Aurora MySQL database cluster.

During peak ingestion events, telemetry processing throughput degrades significantly. Performance monitoring shows that the EC2 instances spend a large percentage of CPU cycles in I/O wait states when writing to the shared file system. Additionally, read queries on the database cluster experience high latency due to replication lag on the read replicas.

Which architectural modifications should a Solutions Architect implement to resolve these performance bottlenecks? (Select TWO.)

  1. Modify the Amazon EFS file system throughput mode to Elastic throughput to dynamically scale storage throughput in response to write activity.Answer
  2. Configure Aurora Auto Scaling to dynamically scale the number of Aurora Replicas based on CPU utilization to distribute the read query workload.Answer
  3. C
    Route read traffic to the database cluster's Multi-AZ standby instance to reduce the read load on the primary writer.
  4. D
    Request AWS Support to pre-warm the Application Load Balancer (ALB) to handle the database replication lag and EFS throughput bottlenecks.
  5. E
    Reduce the Auto Scaling group cooldown period to a duration shorter than the instance bootstrapping time to enable quicker instance launching during spikes.

Answer

The correct options are to modify the Amazon EFS file system throughput mode to Elastic throughput, and to configure Aurora Auto Scaling to dynamically scale the number of Aurora Replicas based on CPU utilization.
The correct options are to change the Amazon EFS throughput mode to Elastic throughput, and to configure Aurora Auto Scaling based on CPU utilization. Elastic throughput ensures that Amazon EFS automatically scales to meet the throughput demands of the instances during peak processing times, eliminating I/O wait states. Configuring Aurora Auto Scaling allows the database cluster to automatically add Aurora Replicas to handle peak read query volumes, which reduces replication lag and latency by distributing the read workload.

Step-by-Step Solution

1
Analyze the file system performance bottleneck where EC2 instances spend time in I/O wait states.
Identify that Provisioned Throughput on Amazon EFS is insufficient for peak write spikes, and transitioning to Elastic throughput will allow automatic scaling.
Elastic throughput dynamically adjusts to write traffic, ensuring tasks do not queue on storage I/O.
2
Analyze the database performance bottleneck where read queries suffer from replication lag on read replicas.
Determine that adding more replicas dynamically via Aurora Auto Scaling will distribute read queries, preventing any single replica from becoming overloaded and falling behind.
Aurora Auto Scaling automatically manages replica count based on target metrics like average CPU utilization.

Key Concept

Resolving storage and database performance bottlenecks by transitioning from static capacity provisioning to dynamic, auto-scaling resource models.
Rate this question

Topics

Optimizing Compute and Storage Performance