Amazon Elastic File System (EFS) Performance

Languages Available: Español (Latinoamérica) | Français | Bahasa Indonesia | Italiano | 日本語 | 한국어 | Português (Brasil) | 中文（简体)

When considering Amazon Elastic File System (EFS), you should understand the performance characteristics of the file system and how Amazon EC2 instance types, I/O size, and parallelism can affect the performance. In this lab, you review scenarios that use various file creation and copy tools to showcase the way that different Amazon Elastic Cloud Compute (EC2) instance types, I/O, and numbers of threads accessing Amazon EFS can have an impact on performance. You also use an Amazon CloudWatch dashboard with custom metrics created using metric math to monitor the activities with common storage metrics.

Level

Advanced

Duration

2 Hours 0 Minutes

Course Objectives

In this course, you will learn how to:

• EC2 instance types affect EFS performance.

• I/O size (block size) and sync frequency affect EFS performance.

• Increasing the number of threads accessing EFS can improve performance.

• Different file transfer tools affect performance when accessing an EFS file system.

• Scale-out data transfer using multiple EC2 instances can improve EFS performance.

Intended Audience

This course is intended for:

• Architects
• Systems Operators

Prerequisites

We recommend that attendees of this course have the following prerequisites:

• Familiar with basic navigation of the AWS Management Console
• Familiar with storage and networking concepts
• Comfortable using Linux commands

Course Outline

• Task 1: How the number of threads used to write data affects Amazon EFS performance

• Task 2: How the network performance of EC2 instance types affect Amazon EFS performance

• Task 3: How I/O size and sync frequency affect throughput to Amazon EFS

• Task 4: How multi-threaded access improves throughput and IOPS

• Task 5: Compare file transfer tools

• Task 6: Using a scale-out architecture with Amazon EFS