Nicolas Buisson
Developer, Principal Cloud Architect for Spikeseed
AWS RE:INVENT 2020 - Week 1
Like every year AWS organizes re:Invent in Las Vegas, a one week event to announce improvements, new features and new services.
Due to the current pandemic, and for the first time, this year it goes all virtual, it’s three long and you can attend for free.
As usual, there are tons of announcements for new services and services updates. Here follows my curated list of major announcements for core services.
Compute
Support of Mac instances for Amazon EC2
AWS has announced the support of macOS on Amazon EC2. It’s a big deal for Apple developers as it will allow to test macOS, iOS, iPadOS, tvOS, and watchOS applications, provisioning environments on demand and based on pay-as-yo-go pricing model.
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/new-use-mac-instances-to-build-test-macos-ios-ipados-tvos-and-watchos-apps/
New EC2 AWS Graviton2 Processors based
AWS has announced new C6gn instances based on their Arm Graviton2 processors. Those instances provide 40% better price/performance compare to current x86-based instances.
Learn more: https://aws.amazon.com/ec2/instance-types/c6/
Lambda - 1ms Billing Granularity Adds Cost Savings
Until now Lambda pricing model was based on number of request and execution time per chunk of 100ms. For now execution time chunk has been reduced to 1ms.
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/new-for-aws-lambda-1ms-billing-granularity-adds-cost-savings/
Lambda – Container Image Support
Among the various runtime (Java, NodeJS, Python, …) available for Lambda, AWS has announced the support of Container Image. In addition AWS has release a Lambda Runtime Interface Emulator enabling developers to test locally they container images by simulating Lambda environment.
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/new-for-aws-lambda-container-image-support/
AWS Fargate support for AWS Batch
AWS has announced the support of AWS Fargate runtime for AWS Batch, meaning AWS Batch jobs could now be executed using serverless computing resources.
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/new-fully-serverless-batch-computing-with-aws-batch-support-for-aws-fargate/
Containers
ECS anywhere and EKS anywhere
It is a very important step forward for hybrid cloud on AWS. Until now hybrid cloud with AWS was based on AWS Outpost, but it is not a solution for any organization. It required to purchase new hardware with a minimal entry price of 100k$ and recurrent monthly fees.
With ECS anywhere and EKS anywhere, AWS brings multi cloud capabilities, allowing to install ECS and EKS outside AWS, on-premises for example. Software stack is deployed on customer managed Virtual Machines using AWS Systems Manager agent. AWS console will be used as a single control plane to manage all clusters, whatever their location. Finally a disconnected mode, meaning without connectivity with the cloud, will be available.
ECS anywhere and EKS anywhere should be available by the first half of 2021.
Dedicated blog post is available here: https://aws.amazon.com/blogs/containers/introducing-amazon-ecs-anywhere/
Amazon EKS Distro
Related to EKS anywhere, AWS has announced Amazon EKS Distro, an open source and free distribution of Kubernetes. EKS Distro provides the same upstream versions of Kubernetes and dependencies, meaning it has been validated operating systems like CentOS, Ubuntu, or RedHat.
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/amazon-eks-distro-the-kubernetes-distribution-used-by-amazon-eks/
New Service - AWS Proton
AWS Proton is a new fully managed service for application deployment based on containers and serverless. It allows to define standard templates for CI/CD pipelines and make them available to your organization.
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/preview-aws-proton-automated-management-for-container-and-serverless-deployments/
Storage
Amazon EBS gp3
Currently Amazon EBS backed by SSD are available in two volume types: io2, replacement of io1 announced in August, provides dedicated/higher IOPS and higher durability (99.999%) , while gp2 provides performances based on volume size.
AWS has announced a new general purpose volume type with gp3. This new volume type offers independent performances of storage capacity, while being 20% cheaper than gp2.
It’s available today, and customers can migrate by simply changing the volume type of their current EBS without any interruption.
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/new-amazon-ebs-gp3-volume-lets-you-provision-performance-separate-from-capacity-and-offers-20-lower-price/
S3 – Strong Read-After-Write Consistency
Until now S3 was considered eventually consistent, meaning there is a delay between a write operation has been successfully executed and it’s result is visible by a read operation.
From now S3 delivers strong read-after-write consistency. It is a really impressive technical improvement especially when considering the CAP Theorem. To summarize S3 provides strong consistency, high availability (99,99%), with high performances (3500 w/s and 5500 r/s per prefix).
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/amazon-s3-update-strong-read-after-write-consistency/
Database
New Service - Babelfish
Over the last years PostgreSQL popularity has increased, a lot. However many customers, for historical reasons, are running proprietary databases like Microsoft SQL Server or Oracle, and after so many investments migration might be difficult.
AWS has announced a new open source, Babelfish for PostgreSQL. It is a Microsoft SQL Server-compatible end-point for PostgreSQL, including T-SQL and TDS protocol support. In other words it allow to run applications written for Microsoft SQL Server without changing libraries, database schema, or SQL statements against PostgreSQL database!
Babelfish is currently available in preview through Amazon Aurora, and will be released in open source in 2021.
Dedicated blog post is available here: https://aws.amazon.com/blogs/opensource/want-more-postgresql-you-just-might-like-babelfish/
Operation
New Service - Amazon DevOps Guru
Amazon DevOps Guru is a new machine learning-powered service allowing to automatically detect operational problems, details the possible causes, and recommends remediation actions. It is based on machine learning models leveraging AWS expertise in running the largest e-commerce website for years. The service collects and analyzes data from various multiple services. Currently services supported are Amazon CloudWatch, AWS Config, AWS CloudTrail, AWS CloudFormation, and AWS X-Ray. Detected anomalies and suggested insights are available through a dashboard while notifications could be configured using SNS.
Dedicated blog post is available here: https://aws.amazon.com/blogs/aws/amazon-devops-guru-machine-learning-powered-service-identifies-application-errors-and-fixes/