AWS Certified SysOps Administrator – Associate — Question 269

A company runs a high performance computing (HPC) application on an Amazon EC2 instance. The company needs to scale this architecture to two or more EC2 instances. The EC2 instances will need to communicate with each other at high speeds with low latency to support the application.

The company wants to ensure that the network performance can support the required communication between the EC2 instances

What should a SysOps administrator do to meet these requirements?

Answer options

• A. Create a cluster placement group. Back up the existing EC2 instance to an Amazon Machine Image (AMI). Restore the EC2 instance from the AMI into the placement group. Launch the additional EC2 instances into the placement group.
• B. Back up the existing EC2 instance to an Amazon Machine Image (AMI). Create a launch template from the existing EC2 instance by specifying the AMI. Create an Auto Scaling group and configure the desired instance count.
• C. Create a Network Load Balancer (NLB) and a target group. Launch the new EC2 instances and register them with the target group. Register the existing EC2 instance with the target group. Pass all application traffic through the NLB.
• D. Back up the existing EC2 instance to an Amazon Machine Image (AMI). Create additional clones of the EC2 instance from the AMI in the same Availability Zone where the existing EC2 instance is located.

Correct answer: A

Explanation

A cluster placement group logically groups instances within a single Availability Zone to enable low-latency, high-throughput network communication, which is ideal for HPC applications. Because an existing EC2 instance cannot be directly moved into a placement group, it must be backed up to an AMI and recreated inside the newly created cluster placement group alongside the new instances. Other methods, such as using an NLB or placing instances in the same Availability Zone without a cluster placement group, do not guarantee the specialized low-latency, high-speed physical network configuration required.