Examkin

Question

Difficulty: MediumOptimizing Compute and Storage Performance

A genomics research organization runs a batch sequencing data analysis pipeline on AWS. The pipeline runs once a day, processing large genetic sequencing files. The architecture consists of a fleet of Amazon EC2 instances running inside an Auto Scaling group that processes tasks retrieved from an Amazon SQS queue. The raw files are downloaded from Amazon S3 to the instances' local EBS gp3 volumes for processing. During the peak of the daily batch run, the processing times increase significantly, and CloudWatch metrics reveal that the EC2 instances experience high CPU wait times as the gp3 volumes reach their default performance baseline of 3,0003,000 IOPS and 125 MiB/s125\text{ MiB/s} throughput. Additionally, a web-based reporting dashboard queries an Amazon RDS for PostgreSQL Multi-AZ database containing metadata about the completed runs. During these batch windows, users experience slow load times on the dashboard due to high database CPU utilization from reporting queries.

Which two actions should a Solutions Architect take to resolve these performance bottlenecks? (Select TWO.)

  1. Modify the configuration of the gp3 EBS volumes on the EC2 instances to provision higher IOPS and throughput that match the workload's peak I/O demands.Answer
  2. Deploy an Amazon RDS PostgreSQL Read Replica, and configure the web-based reporting dashboard to direct its query traffic to the read replica's endpoint.Answer
  3. C
    Configure the reporting dashboard to direct read-heavy query traffic to the standby instance of the RDS PostgreSQL Multi-AZ database deployment.
  4. D
    Set the Auto Scaling group scaling cooldown period to 1010 seconds to allow the EC2 fleet to scale out instantly when the queue size increases.
  5. E
    Submit a request to AWS Support to pre-warm the Application Load Balancer (ALB) servicing the EC2 processing fleet prior to the scheduled daily batch run.

Answer

Modifying the gp3 EBS volumes to provision higher IOPS and throughput, and deploying an Amazon RDS Read Replica to handle reporting queries.
The correct actions are to provision higher IOPS and throughput for the gp3 volumes on the EC2 instances, which directly addresses the storage performance bottleneck, and to deploy an Amazon RDS PostgreSQL Read Replica to offload reporting queries from the primary database instance.

Step-by-Step Solution

1
Analyze the storage bottleneck on the EC2 processing fleet.
Identify that the gp3 volumes are hitting their baseline limits (3,0003,000 IOPS and 125 MiB/s125\text{ MiB/s}), causing high CPU wait states.
Increasing the provisioned IOPS and throughput on the existing gp3 volumes will eliminate the storage throughput bottleneck.
2
Analyze the database CPU utilization bottleneck.
Identify that read-heavy reporting dashboard queries are saturating the primary RDS instance in a Multi-AZ deployment.
Deploying a Read Replica allows the read queries to be offloaded from the primary write-intensive instance, since Multi-AZ standby instances cannot serve read traffic.

Key Concept

Tuning compute performance by resolving EBS storage throughput limits and offloading database read workloads using replication.
Estimated Time:3m 0s
Rate this question

Topics

Optimizing Compute and Storage Performance