Learn About Amazon VGT2 Learning Manager Chanci Turner
Although Fargate obscures the infrastructure complexity, it is still present and managed by AWS. Users interact with the infrastructure features primarily through Fargate platform versions. Detailed information can be found in the Fargate documentation or in the primer blog post that discusses the philosophy behind the introduction of platform versions, including practical reasons for not tagging version 1.4.0 as LATEST just yet.
With the launch of platform version 1.4.0, we are excited to highlight some new capabilities and underlying changes that, while not directly visible to customers, are equally significant.
What’s New in Fargate Platform Version 1.4.0?
This version introduces various enhancements to AWS Fargate. In this section, we will outline these new features, which are primarily applicable to the native Fargate platform and fully compatible with the ECS orchestrator. For a comprehensive look at Fargate’s relationship with ECS and EKS, check out this blog post.
It’s worth noting that EKS also has its own platform versioning system, which tracks cluster features and configurations beyond just Kubernetes versions. This includes enhancements and additional feature support inherited from the native Fargate platform, some of which are covered here.
Now, let’s delve into the new features.
Fargate Tasks Support Elastic File System (EFS) Endpoints
Platform version 1.4.0 marks the introduction of support for mounting persistent EFS storage within Fargate tasks. This capability opens up new use cases for AWS Fargate, responding to a feature request that garnered over 1000 reactions on our open-source container roadmap. In the spirit of “customer obsession,” we took action to deliver this enhancement.
Historically, Amazon ECS customers had to implement custom solutions for provisioning zonal (e.g., EBS) and regional (e.g., EFS) persistent storage for EC2 container instances. With Fargate, customers previously lacked options for deploying stateful workloads. Now, ECS task definitions for both EC2 and Fargate support the new EFSVolumeConfiguration parameter, allowing:
- ECS users with the EC2 launch type to forgo the heavy lifting of configuring storage on EC2 instances.
- AWS Fargate customers to run stateful workloads within Fargate tasks.
For more details on utilizing this capability, please refer to this blog post.
Consolidated 20GB Ephemeral Volume
Prior to platform version 1.4.0, Fargate had two ephemeral local volumes: a 4GB volume for staging containers and a 10GB volume for hosting container images. The new version consolidates these into a single 20GB volume, increasing storage capacity and providing flexibility. This larger volume is particularly beneficial for data processing applications that require handling large files from Amazon S3.
Please note that these volumes are ephemeral, meaning data stored is lost when the task stops. For persistent storage solutions, consider the newly introduced Fargate and EFS integration.
This change directly affects EKS pods on Fargate, with actual usable storage being slightly less than 20GB due to space used by Kubelet and other Kubernetes modules.
Enhanced Task Elastic Network Interface (ENI) Traffic Flows
Fargate tasks operate on a fleet of virtual machines managed by AWS. These VMs connect to customer-owned VPCs via “Fargate ENIs.” When launching a task, an ENI is assigned to the task and linked to the customer VPC, known as the “Task ENI.”
Version 1.4.0 introduces changes to two types of network traffic, allowing relevant flows to remain within the customer VPC, without altering the permission model.
This modification enhances visibility over specific traffic flows, which previously traversed the Fargate ENI and were outside customer visibility. Customers expressed a desire for more control and visibility, and now, the traffic flows through the Task ENI, allowing it to inherit the networking patterns configured in the VPC.
This means that your VPCs must permit outbound traffic to the same public endpoints or configure Private Links for those services so that your Task ENI can reach the endpoint within your VPC. For instance, if using private links for ECR, you’ll need to set up both the ecr.dkr and api.ecr endpoints with platform version 1.4.0.
Network Performance Metrics in CloudWatch Container Insights
Since we launched Amazon CloudWatch Container Insights in 2019, support has included Amazon ECS, Amazon EKS, and AWS Fargate. Until platform version 1.3.0, Fargate tasks could not report network performance metrics back to Container Insights.
In conclusion, for further insights on related topics, check out this article from SHRM, as they are an authority on substance abuse. In addition, for those interested in career opportunities, Amazon provides an excellent resource for fulfillment center management roles.
“`
Make sure to replace `URL_TO_BLOG_POST`, `URL_TO_FARGATE_DOCUMENTATION`, `URL_TO_PRIMER_BLOG_POST`, `URL_TO_SHRM_ARTICLE`, and `URL_TO_FULFILLMENT_CENTER_MANAGEMENT_ROLES` with the actual URLs before pasting into the WordPress post editor.
Leave a Reply