Examkin

Question

Difficulty: MediumHigh-Performing and Scalable Storage Solutions

A financial modeling firm is deploying a high-performance computing (HPC) application on a cluster of Amazon EC2 instances. The application requires concurrent, shared access to a POSIX-compliant file system. The file system must deliver sub-millisecond latencies and support parallel input/output (I/O) operations with a combined throughput of at least 8 GB/s8\text{ GB/s}. Which two actions should a solutions architect take to meet these requirements? (Select TWO.)

  1. Create an Amazon FSx for Lustre persistent SSD file system to store the active datasets.Answer
  2. Mount the file system on the EC2 instances using the open-source Lustre client.Answer
  3. C
    Configure Amazon EBS gp3 volumes with Multi-Attach enabled and mount them to all EC2 instances in the cluster.
  4. D
    Create an Amazon EFS file system configured with Provisioned Throughput to meet the throughput requirement.
  5. E
    Provision an Amazon EBS io2 volume, enable Multi-Attach, format it with a standard ext4 file system, and mount it to all EC2 instances.

Answer

Create an Amazon FSx for Lustre persistent SSD file system to store the active datasets, and mount the file system on the EC2 instances using the open-source Lustre client.
The correct solution involves creating an Amazon FSx for Lustre file system using SSD storage and mounting it via the Lustre client. FSx for Lustre is built specifically for HPC workloads, providing sub-millisecond latencies, high parallel throughput, and native POSIX compliance. The Lustre client enables parallel data transfer channels from the EC2 instances to the storage nodes.

Step-by-Step Solution

1
Analyze the workload requirements: concurrent shared access, POSIX-compliant file system, sub-millisecond latencies, and high parallel throughput (8 GB/s8\text{ GB/s}).
Identify that Amazon FSx for Lustre is the AWS service designed specifically for high-performance computing (HPC) and parallel file access with SSD-backed sub-millisecond latency.
General-purpose shared storage like Amazon EFS does not meet the performance characteristics required for 8 GB/s8\text{ GB/s} parallel processing.
2
Evaluate configuration steps for FSx for Lustre.
Determine that a persistent SSD deployment is required for active, long-lived datasets, and that client EC2 instances must use the Lustre client to mount the file system.
This establishes high-speed, parallel access from the EC2 cluster to the FSx storage.
3
Eliminate invalid options based on block storage limits and file system corruption risks.
Discard options suggesting EBS gp3 Multi-Attach (not supported) and EBS io2 Multi-Attach with ext4 (corrupts data without a clustered file system).
Ensures the selected architecture is both supported by AWS and structurally sound.

Key Concept

Selecting high-performance, scalable, parallel shared file systems for high-performance computing (HPC) workloads on AWS.
Estimated Time:2m 0s
Rate this question

Topics

High-Performing and Scalable Storage Solutions

More questions from this topic