New – Announcing HAQM EFS Elastic Throughput

Today, we are announcing the availability of HAQM EFS Elastic Throughput, a new throughput mode for HAQM EFS that is designed to provide your applications with as much throughput as they need with pay-as-you-use pricing. This new throughput mode enables you to further simplify running workloads and applications on AWS by providing shared file storage that doesn’t need provisioning or capacity management.

Elastic Throughput is ideal for spiky and unpredictable workloads with performance requirements that are difficult to forecast. When you enable Elastic Throughput on an HAQM EFS file system, you no longer need to think about actively managing your file system performance or over-paying for idle resources in order to ensure performance for your applications. When you enable Elastic Throughput, you don’t specify or provision throughput capacity, HAQM EFS automatically delivers the throughput performance your application needs while you the builder pays only for the amount of data read or written.

HAQM EFS is built to provide serverless, fully elastic file storage that lets you share file data for your cloud-based applications without having to think about provisioning or managing storage capacity and performance. With Elastic Throughput, HAQM EFS now extends its simplicity and elasticity to performance, enabling you to run an even broader range of file workloads on HAQM EFS. HAQM EFS is well suited to support a broad spectrum of use cases that include analytics and data science, machine learning, CI/CD tools, content management and web serving, and SaaS applications.

A Quick Review
As you may already know, HAQM EFS already has the Bursting Throughput mode, which is available as a default and supports bursting to higher levels for up to 12 hours a day. If your application is throughput constrained on Bursting mode (for example, utilizes more than 80 percent of permitted throughput or exhausts burst credits), then you should consider using Provisioned (which we announced in 2018), or the new Elastic Throughput modes.

With this announcement of Elastic Throughput mode, and in addition to the already existing Provisioned Throughput mode, HAQM EFS now offers two options for workloads that require higher levels of throughput performance. You should use Provisioned Throughput if you know your workload’s performance requirements and you expect your workload to consume a higher share (more than 5 percent on average) of your application’s peak throughput capacity. You should use Elastic Throughput if you don’t know your application’s throughput or your application is very spiky.

To access Elastic Throughput mode (or any of the Throughput modes), select Customize (selecting Create instead will create your file system with the default Bursting mode).

Create File system

New – Elastic Throughput

You can also enable Elastic Throughput for new and existing General Purpose file systems using the HAQM EFS console or programmatically using the HAQM EFS CLI, HAQM EFS API, or AWS CloudFormation.

Elastic Throughput in Action
Once you have enabled Elastic Throughput mode, you will be able to monitor your cost and throughput usage using HAQM CloudWatch and set alerts on unplanned throughput charges using AWS Budgets.

I have a test file system elasticblog that I created previously using the HAQM EFS console, and now I cannot wait to see Elastic Throughput in action.

File system (elasticblog)

I have provisioned an HAQM Elastic Compute Cloud (HAQM C2) instance which I mounted to my file system. This EC2 instance has data that I will add to the file system.

I have also created CloudWatch Alarms, which will monitor throughput usage and set alarm thresholds (ReadIOBytes, WriteIOBytes, TotalIOBytes, and MetadataIOBytes).

CloudWatch for Throughput Usage

The CloudWatch dashboard for my test file system elasticblog looks like this.

CloudWatch Dashboard – TotalIOBytes for File System

Elastic Throughput allows you to drive throughput up to a limit of 3 GiB/s for read operations and 1 GiB/s for write operations per file system in all Regions.

Available Now
HAQM EFS Elastic Throughput is available in all Regions supporting EFS except for the AWS China Regions.

To learn more, see the HAQM EFS User Guide. Please send feedback to AWS re:Post for HAQM Elastic File System or through your usual AWS support contacts.

– Veliswa x