You can now create Amazon Managed Streaming for Apache Kafka (Amazon MSK) clusters for less than $2.50/day using kafka.t3.small brokers.  

Amazon Elastic Container Service (ECS) tasks running on both Amazon Elastic Compute Cloud (EC2) and AWS Fargate can now mount Amazon Elastic File System (EFS) file systems. ECS tasks using EFS will automatically mount the file systems specified by the customer in the task definition and make them available to the containers in the task across all availability zones in the region. This enables persistent, shared storage to be defined and used at the task and container level in ECS. 

Amazon RDS Proxy, a fully managed, highly available database proxy for Amazon Relational Database Service (RDS), is now available in preview for Aurora PostgreSQL and RDS PostgreSQL. RDS Proxy makes applications more scalable, more resilient to database failures, and more secure.

You can now use the AWS Elastic Beanstalk ListPlatformBranches API action that lists available Elastic Beanstalk platform branches and provides summary information about each platform branch. You can use the API action to query the platform state (Supported, Beta, Deprecated, Retired) of your Elastic Beanstalk environment. For more information on Platform Branch state, see Elastic Beanstalk Platforms Glossary in the AWS Elastic Beanstalk Developer Guide. For details about the ListPlatformBranches API action, see ListPlatformBranches in the AWS Elastic Beanstalk API Reference. See Release Notes for additional details.

You can now run your applications on AWS Elastic Beanstalk using the new generation of Docker, Corretto and Python Elastic Beanstalk platforms built on top of Amazon Linux 2 Operating System.

AWS Service Catalog now supports local launch constraints. With this new feature, administrators can share products from one AWS account to other AWS accounts, and have those products launch locally in each account. In Service Catalog, this is referred to as a hub-and-spoke model of sharing approved IT services. This makes it easier for administrators to distribute and manage these services across multiple accounts, providing self-service access while staying compliant with organizational best practices and policies.

Amazon RDS for Oracle now supports 19.2 version of Oracle Application Express (APEX) for 11.2, 12.1, 12.2, 18c and 19c versions of Oracle Database. Using APEX, developers can build applications entirely within their web browser. 

VPC Interface endpoints (powered by AWS PrivateLink) and gateway endpoints now support additional IAM condition keys. With this launch, you can now use the IAM condition key ec2: VpceServiceOwner to restrict creation of interface endpoints to either AWS services or services owned by specific AWS accounts. You can also restrict endpoint creation to only specific services by using the ec2: VpceServiceName condition key in your IAM policies. Additionally, you can manage actions on your VPC endpoint and endpoints services based on existing tags on resources like VPCs and subnets using the ec2:resourceTag condition key. 

Amazon DynamoDB global tables are now available in the AWS China (Beijing) Region, operated by Sinnet, and the AWS China (Ningxia) Region, operated by NWCD. With DynamoDB global tables, you can create fully replicated tables across regions for disaster recovery and high availability of your DynamoDB tables. With this launch, you can now add a replica table in one AWS China Region to your existing DynamoDB table in the other AWS China Region. When you use DynamoDB global tables, you benefit from an enhanced 99.999% availability SLA at no additional cost.  

In the past, when submitting an encoding job to an AWS Elemental MediaConvert reserved queue, the job would stay in that queue until it could be processed. If the queue was busy, the job would wait to start until previous jobs completed. This presents a challenge for videos requiring strict turnaround times. To solve this, MediaConvert now has the ability to submit jobs using a feature called queue hopping. You specify the queue a job should start in and how long it should wait before hopping to a different queue. With queue hopping, you can avoid long wait times for transcoding jobs that are time-sensitive while also effectively utilizing your reserved transcoding slots. For example, a job can be submitted to a reserved queue with a wait time of 15 minutes; if the job has not started after that time, it can hop to an on-demand queue to start encoding. For more information about queue hopping, please read the documentation .