Discuss Amazon SAP-C02 Exam Topic 1 Question 70

A company has built a high performance computing (HPC) cluster in AWS tor a tightly coupled workload that generates a large number of shared files stored in Amazon EFS. The cluster was performing well when the number of Amazon EC2 instances in the cluster was 100. However, when the company increased the cluster size to 1,000 EC2 instances, overall performance was well below expectations.

Which collection of design choices should a solutions architect make to achieve the maximum performance from the HPC cluster? (Select THREE.)

AEnsure the HPC cluster Is launched within a single Availability Zone.

BLaunch the EC2 instances and attach elastic network interfaces in multiples of four.

CSelect EC2 Instance types with an Elastic Fabric Adapter (EFA) enabled.

DEnsure the cluster Is launched across multiple Availability Zones.

EReplace Amazon EFS with multiple Amazon EBS volumes in a RAID array.

FReplace Amazon EFS with Amazon FSx for Lustre.

Suggested Answer: A, C, F

A . High performance computing (HPC) workload cluster should be in a single AZ.

C . Elastic Fabric Adapter (EFA) is a network device that you can attach to your Amazon EC2 instances to accelerate High Performance Computing (HPC)

F . Amazon FSx for Lustre - Use it for workloads where speed matters, such as machine learning, high performance computing (HPC), video processing, and financial modeling.

Cluster -- packs instances close together inside an Availability Zone. This strategy enables workloads to achieve the low-latency network performance necessary for tightly-coupled node-to-node communication that is typical of HPC applications.

https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/placement-groups.html

Get Premium SAP-C02 Questions as Interactive Web-Based Practice Test or PDF Real Exam Format Detailed Explanations Instant Access Get Your Practice Test

20 comments · newest first

Jian Pham 5 days ago

Maybe A, C, F is not correct because the tricky part is that latency within a single AZ matters most.

Celestine 22 days ago

EFA really boosts network performance.

Garry 27 days ago

RAID with EBS could help, but not as much.

Serina 1 month ago

A single AZ won't cut it for 1,000 instances.

Dudley 1 month ago

Agreed, those options maximize throughput.

Art 1 month ago

I’d go with C, D, and F.

Celestine 2 months ago

FSx for Lustre seems like a solid choice.

Garry 2 months ago

EFS can't handle that many instances.

Serina 2 months ago

Definitely need to consider multiple AZs.

Dudley 2 months ago

I think EFA is key for performance.

Art 2 months ago

This question is tricky.

Jose 2 months ago

Launching in a single AZ could limit performance, right?

Denna 3 months ago

Really? I thought EFS was fine for large workloads.

Alex 3 months ago

FSx for Lustre is way faster than EFS, no doubt!

Shelba 3 months ago

I think using multiple AZs is overkill for this.

Krissy 3 months ago

EFA is a must for HPC workloads!

Serina 3 months ago

I recall something about network interfaces, but I can't remember if launching them in multiples of four is actually beneficial. I might be leaning towards option B, but I'm not confident.

Jolanda 3 months ago

I practiced a similar question where EFS was a bottleneck, and I think replacing it with FSx for Lustre could really improve performance, so I’d go with option F.

Olga 4 months ago

I'm not entirely sure, but I feel like launching across multiple Availability Zones could help with performance, so maybe option D is worth considering?

Nieves 4 months ago

I remember studying about EFA and its benefits for HPC workloads, so I think option C is definitely a good choice.

Amazon SAP-C02 Exam - Topic 1 Question 70 Discussion