Kinesis Streaming Tables Can Cut Iceberg Delivery and Query Costs

Amazon Kinesis Data Streams now offers serverless streaming tables for Apache Iceberg tables on Amazon S3 Tables, including inline compaction.

AWS states that the feature can reduce delivery costs by up to 50% and downstream query costs by up to 30%. It does this by removing the need for custom delivery pipelines and reducing small-file overhead.

AWS Agent Registry Adds Tags for Cost Allocation and Governance

AWS Agent Registry is now generally available with tagging for registries and records, supporting organization, cost allocation, and access-control use cases.

The service also supports organization-wide sharing through AWS Resource Access Manager and infrastructure-as-code provisioning. These capabilities can help FinOps and platform teams govern growing collections of agents.

Lambda Recursive Loop Detection Helps Prevent Unexpected Usage

Lambda recursive loop detection is now available in all commercial AWS Regions and automatically stops supported runaway invocation loops.

By preventing accidental recursive workloads from generating unbounded executions, the feature helps avoid unexpected usage and billing. This provides a direct safeguard against a specific failure mode that could otherwise continue consuming compute.

CloudWatch Agent Filtering Can Help Control Journald Log Costs

The CloudWatch agent can now collect systemd journald logs directly and filter them by units, priorities, fields, and regular expressions.

Native filtering can reduce unnecessary log volume before it reaches CloudWatch Logs. That can help control ingestion costs, particularly across large AWS environments.

OpenSearch Cluster Insights Helps Target Capacity Changes

Amazon OpenSearch Service has added 17 Cluster Insights to identify resource exhaustion and configuration issues.

The insights cover conditions such as sustained central processing unit saturation, Java virtual machine memory pressure, zone imbalance, and replica misconfiguration. The no-additional-cost recommendations can guide targeted scaling and capacity changes across 11 AWS Regions.

EC2 R9g and R9gd Graviton5 Instances Reach General Availability

Amazon EC2 R9g and R9gd memory-optimized instances powered by Graviton5 are now generally available.

AWS states that the instances offer up to 25% better compute performance than Graviton4-based predecessors and up to 30% faster database performance. They’re available through On-Demand, Spot, Dedicated Instances, Dedicated Hosts, and Savings Plans.