Every year, since 2012, a November is the most awaited time of the year for the AWS community. It’s the time of keynote announcements, training and certification opportunities. It’s the time of Amazon re:Invent conference, hosted by Amazon Web Services in Las Vegas for the global cloud computing community.
If you are an engineer, developer, sysop, solution architect or whoever interested in public cloud solutions, it’s something worth your interest. If you are not able to attend it in person, you can watch the event online. However, you will not be able to take part in the after-hours event and much more. Anyway, this year re:Invent is over. Before you start booking your flights and hotel for re:Invent 2018, let’s take a look at the biggest news of re:Invent 2017.
Because there were a lot of announcements from Amazon, we took 10 of them we consider the most important. Here is the list.
EC2 Bare Metal
Amazon EC2 Bare Metal instances are the new instance types which you can use not only like regular ones, with the full support of different AWS services like Virtual Private Cloud, Elastic Block Storage, Elastic Load Balancing, Auto Scaling and so on, but also you can take advantage of full hardware access. This means direct access to the processor an memory of the underlying server. With Bare Metal instances you can leverage such low-level hardware features as Intel VT, and you can run applications that need to be run in the non-virtualized environment. If you have such applications, which might be a showstopper for moving to the cloud, they are not anymore. You don’t have to keep your old bare metal servers on premises, and you can migrate them to Amazon.
EC2 Bare Metal instances are available for preview in the form of the i3.metal instance type. To get access to them, you need to request it first on the AWS web page.
Amazon FreeRTOS is an operating system for microcontrollers. It makes them easy to program, deploy, secure, connect and manage. It is based on FreeRTOS open source kernel, which means to be the most popular, real-time operating system for microcontrollers, and if you already use it on your devices, with Amazon FreeRTOS, you can take advantage of the functionality provided by additional libraries. For instance, it includes support for data encryption, key management or network connectivity with Wi-Fi or Ethernet. Devices using Amazon FreeRTOS can connect directly to AWS IoT or AWS Greengrass, so you can easily connect data from your devices and use it for your IoT applications.
Amazon FreeRTOS is free of charge, open source and available to all today.
S3 Select and Glacier Select
These are two different services, both related to S3. Amazon S3 Select allows you to offload the heavy lifting of filtering and accessing data inside the object to S3 service. Instead of retrieving the entire object to process, you can retrieve only a subset of data, selected by SQL expression. This means, that when you store a large .csv file in S3, you don’t have to download and process the entire file to get the data you need. You can just use a simple SQL query to get the data you are interested in, and nothing else. What’s even better, you can use a Presto connector to use S3 Select with Amazon EMR. Using S3 Select can introduce a few times performance improvement and significant cost reduction.
Glacier Select is similar to S3 Select. It allows you to run SQL queries directly against data archived in Glacier and retrieve only the data you need to use for analytics. It makes easy to find and retrieve from archive records matching for instance only specific customer.
Amazon Glacier Select is generally available today in all AWS regions where Glacier is offered. S3 Select is offered as a preview in selected regions.
Amazon Aurora Serverless
It’s one of the top announcements from re:Invent. Aurora Serverless is an on-demand and auto-scaling configuration for Amazon Aurora. It can start up, shut down and scale up and down your database adjusting to the application needs. You no longer need to predict how powerful your database instance should be and you don’t have to manage any database instance or cluster anymore. This makes Aurora Serverless a perfect solution for applications with infrequent, intermittent or unpredictable workloads with no worries about over- or under-provisioning of resources. Aurora Serverless is available as a preview for the Aurora MySQL-compatible edition, and you have to signup if you want to participate.
When talking about Aurora, it is also worth to mention about Aurora Multi-Master preview. This feature allows Aurora to create multiple write nodes, scaling both, reads and writes across multiple availability zones. Same as for Aurora Serverless, this preview is available for the Aurora MySQL-compatible edition.
Amazon Elastic Container Service for Kubernetes
If you run Kubernetes on AWS, this is a service you will like. Amazon Elastic Container Service for Kubernetes (Amazon EKS) is a managed service that makes easy to run Kubernetes on AWS without needing to install, operate and maintain own Kubernetes clusters. It is fully compatible with applications running on the standard Kubernetes environment, so you can easily migrate your Kubernetes applications to EKS. Amazon EKS runs three Kubernetes masters across three Availability zones to ensure high-availability and it also automatically replaces unhealthy masters. EKS is integrated with other AWS services and is available as a preview.
AWS Fargate is a technology for Amazon ECS and EKS that allows you to run containers with no need to provision, configure and scale clusters of virtual machines. It similar to EC2, but instead of the virtual instance, it gives you a container which is still fully integrated with AWS. With AWS Fargate you can focus on building your applications and not on managing your infrastructure.
AWS Fargate is available on Amazon ECS. Support for Amazon EKS will be available in 2018. So far it is available in US East (N. Virginia) region only, but other regions are to come soon.
Alexa for Business
Alexa for Business introduces well-known Alexa to organizations providing the tools and controls to use it on an organization scale. Alexa can help people to manage calendars, dial into meetings or search for information with natural language. Alexa for Business lets you use your existing conference room hardware and meeting applications. It also provides an API to allow Alexa to work with additional equipment or perform specific tasks. It is pay-as-you-go service with no up-front fees and no long-term commitments. Alexa for Business is available in the US East (N. Virginia) region.
Amazon Rekognition Video
It is a deep learning powered video analysis service. It can detect activities, track and recognize people, objects, etc., including face recognition in live streams from Amazon Kinesis Video Streams. Rekognition Video returns a bounding box with the specific location of the person or face in the frame. It is available in the US East (N. Virginia), US West (Oregon) and EU (Ireland) regions.
SageMaker is like machine learning for everyone. It takes away heavy lifting of machine learning and allows to build, train and deploy machine learning models quickly and easily. It is a fully-managed service, so you don’t have to worry about underlying hardware.
SageMaker includes the most common machine learning algorithms, it is preconfigured to run TensorFlow and Apache MXNet (you can use your own framework as well), and it can tune your model to achieve the highest possible accuracy.
SageMaker is available in the US East (N. Virginia and Ohio), US West (Oregon) and EU (Ireland) regions.
Inter-Region VPC Peering
This feature allows you to spread your AWS infrastructure across different AWS regions. With Inter-Region VPC Peering VPC resources like EC2 instances, RDS databases and Lambda functions running in different regions are allowed to communicate with each other using private IP addresses. No extra gateways, network appliances or VPN connections are required. It is a simple and cost-effective way to share resources between regions. Data transferred across regions is encrypted, and there is no single point of failure or bandwidth bottleneck, as inter-region traffic always stays on the global AWS backbone and never traverses the public internet.
Data transfer across regions is charged at the standard inter-region data transfer rates. Inter-Region VPC Peering is available in the US East (N. Virginia and Ohio), US West (Oregon) and EU (Ireland) regions. Other regions are to come soon.
There were over sixty new features and services announced during this year re:Invent. We’ve chosen ten which we believe are most important. However, you can consider other, not mentioned here, more important for you. AWS is growing up almost every day. New features and services are announced not only during the conferences. Each month comes with some news in different categories. We advise you to follow the ones that are most important to you. Some of new features or service may be available in preview only or in regions which are not preferred by you, but soon they should be generally available in more regions, so you need to keep your finger on the pulse. So, don’t forget to register for the re:Invent 2018. See you in Las Vegas!