AWS Elastic File System (EFS) Summary

Michael Weeks
Absolute Zero
Published in
4 min readNov 21, 2019

--

This will be another high-level overview of an AWS service, specifically EFS. The goal here is to immerse you in some of the building blocks so that you can pass your Certified Solutions Architect Associate (CSAA) exam.

Overview

Amazon Elastic File System (EFS) is a scalable cloud file storage solution for use with EC2 instances. It’s elastic because it will automatically grow and shrink as you add/remove files. It has a simple interface that enables you to create and configure file systems quickly and simply. It’s similar to EBS, but with EBS you can only mount your virtual disk to one EC2 instance. You can have 2 instances sharing an EFS volume.

Key Features:

  • Supports NFSv4 Protocol* (specifically v4.0 and v4.1)
  • Allows for multiple EC2 instance access at the same time
  • Common data source for workloads and apps running on more than one instance or server
  • Offers file access semantics such as strong data consistency and file locking
  • Highly scalable (can grow to PB scale), highly available, and highly durable
  • Allows for high levels of throughput
  • Offers two storage classes: Standard and Infrequent Access
  • You only pay for storage used by your file system
  • Control access to file systems using Portable Operating System Interface (POSIX) permissions
  • EFS-to-EFS backup system allows you to automatically schedule incremental backups of your EFS file system
  • AWS DataSync — Allows you to move and manage EFS file data that makes it faster and easier to move data between on-prem storage and EFS

How it Works

EFS provides file storage within the cloud. You create the EFS file system, mount it on an EC2 instance, and then you’re able to read/write data to/from your system. It’s also available in your VPC through NFSv4 protocol.

Make sure to select an AMI that supports the NFSv4 protocol when setting up your EC2 instance (Linux current gen NFSv4.1 client recommended). For select AMIs, you’ll need to actually install a NFS client to mount your file system on your instance. EFS allows concurrent access from EC2 instances in your VPC allowing apps to scale beyond a single connection.

  • You can mount EFS on instances in only one VPC at a time

EFS uses mount targets which allows you to access EFS in a VPC. This provides an IP address for a NFSv4 endpoint where you can mount EFS. You mount it by using its DNS name. This resolves to the IP address of the EFS mount target in the same availability zone as your EC2 instance.

You can create one mount target in each availability zone in an AWS region. If there are multiple subnets in an availability zone in your VPC, you create a mount target in one of the available subnets. All EC2 instances in that AZ will share that mount target from there.

EFS can be mounted on an EC2 instance or on-premises through an AWS Direct Connect connection.

EFS Documentation

The above architecture is an example of how mount targets should be implemented in each availability zone.

EFS Performance Modes

EFS performance modes come in two different flavors. You select a performance mode when you create one. These come at no additional cost, EFS is billed the same regardless of modes.

General Purpose Performance Mode

  • This is the default mode and it’s best used for latency-sensitive use cases.
  • Use cases include web serving environments, content management systems, home directories, and general file serving.

Max I/O Performance Mode

  • This is used to scale to higher levels of aggregate throughput and operations per second. The trade-off is slightly higher latencies for file operations.
  • Use cases include highly parallelized apps and workloads like big data analysis, genomics analysis, and media processing.

Throughput Modes

You get two flavors of throughput modes for EFS:

Bursting Throughput

  • This is the default mode
  • Throughput on EFS scales as your file system grows
  • File-based workloads are typically spiky, hence ‘bursting’ throughput to adequately accommodate the spikes

Provisioned Throughput

  • This mode is where you specify the throughput of your file system independent of the amount of data stored

Limitations

EFS is limited to 1,000 file systems per AWS Region. You can submit a request to Amazon to increase this limit.

In Conclusion

This, like all of my AWS articles, isn’t a comprehensive guide but a snapshot of what they do at a high level. EFS is pretty cool in the sense that it can be mounted on more than one EC2 instance at the same time which allows access to files on EFS at the same time. It’s roughly 10x more expensive than EBS but it also gives you extra features.

--

--