AWS Updates AgentCore and S3 Tables to Cut Migration and Runtime Costs

Amazon Web Services has introduced a series of updates designed to speed up data migrations and reduce operational overhead for cloud builders. The recent announcements focus on minimizing system downtime and optimizing resource usage across several core cloud services.

Streamlined Data Migration and Bulk Loads

Moving large datasets to modern formats typically requires weeks of custom engineering work. A new feature for Amazon S3 Tables allows developers to migrate Apache Hive tables registered in the AWS Glue Data Catalog directly to Apache Iceberg. The migration tool validates the source data and synchronizes only the changed records, meaning production systems can remain operational during the process. Similarly, Amazon DynamoDB now supports bulk item loading directly from S3 into existing tables using managed, asynchronous jobs that utilize dedicated system capacity.

Faster AI Agents and New Runtime Features

For developers building artificial intelligence applications, the platform has released the Amazon Bedrock AgentCore Runtime V2. This updated runtime improves memory management by starting sessions small and paging in data on demand, releasing memory when it is no longer active. It also tackles cold starts by preparing and snapshotting the environment beforehand, keeping start times consistent even as container image sizes grow.

• AgentCore Runtime V2 reduces peak memory holding and maintains fast start times.
• S3 Tables enables managed migration from Hive to Apache Iceberg without stopping writes.
• DynamoDB now supports managed, asynchronous bulk data loading from Amazon S3.

Infrastructure and Security Enhancements

Other updates include a new Web Answer tool for AgentCore that returns cited, synthesized answers to natural-language questions, and a queue mode for self-managed Apache Kafka event sources in AWS Lambda. Additionally, AWS Security Hub now provides a centralized inventory of software bills of materials for compute resources, aggregating package versions across EC2 instances, ECR images, and Lambda functions.