SOA-C02 Question 337
Select 3Your organization is running a high-performance computing application that requires low-latency networking and high throughput between EC2 instances. To optimize the performance, you want to configure the instances to maximize network performance. Which of the following configurations should you implement?
- A
Launch the EC2 instances in a cluster placement group.
- B
Use Elastic Network Adapter (ENA) on the EC2 instances.
- C
Attach Amazon EBS volumes to the EC2 instances for high-speed local storage.
- D
Use instance types that support enhanced networking.
- E
Launch the instances across multiple Availability Zones to improve network speed.
Show answer and explanation
Correct answers: A, B, D
Explanation
To optimize network performance for high-performance computing, you should launch the EC2 instances in a cluster placement group, use Elastic Network Adapter (ENA), and select instance types that support enhanced networking. These configurations collectively reduce latency and increase throughput between the EC2 instances, which is essential for your application. While EBS volumes and multi-AZ deployments have their use cases, they do not directly address the network performance optimization required in this scenario.
- A. Correct.
Launching EC2 instances in a cluster placement group improves network performance by grouping instances closely together in the same Availability Zone, allowing for low-latency and high-throughput communication.
- B. Correct.
Elastic Network Adapter (ENA) enables enhanced networking, which is critical for high-performance applications requiring low-latency and high throughput.
- C. Incorrect.
Amazon EBS volumes are network-attached and do not provide the same level of performance as instance store volumes for local, high-speed storage. This option does not address network performance.
- D. Correct.
Certain instance types are optimized for enhanced networking, which provides better throughput and lower latency. Selecting such instance types ensures better network performance.
- E. Incorrect.
Launching instances across multiple Availability Zones might improve fault tolerance but will not enhance network speed between the instances as inter-AZ communication introduces higher latency.