As a Cloud Security Architect, DevSecOps Lead, and Terraform Specialist, I'm pleased to present today's comprehensive summary of new AWS features and service updates, focusing on their security implications and Infrastructure as Code (IaC) readiness.
Database Highlights
- Amazon RDS for MySQL: The new MySQL 26.10 preview environment introduces critical bug fixes, security patches, and significant performance enhancements for
READ COMMITTEDworkloads, alongside improved purge lag warnings. Preview instances are ephemeral (max 60 days) and snapshots are restricted to the preview environment, ensuring isolation and reducing risk to production data. From a security and architectural standpoint, the preview environment offers a safe sandbox for testing updates without impacting production, while performance improvements contribute to overall system stability and resilience. *Terraform Support*: Supported in Terraform AWS Provider: aws_db_instance (argumentengine_version) in v5.0.0+. - Amazon RDS for PostgreSQL: AWS has announced Extended Support minor versions (13.23, 12.22, 11.22) for PostgreSQL, providing continued security and bug-fix support beyond standard end-of-life dates. This allows organizations more flexibility in planning major version upgrades while maintaining a secure and stable database environment. Upgrades can be performed using Blue/Green Deployments, in-place upgrades, or snapshot restores, offering robust options for minimizing downtime and risk. *Terraform Support*: Supported in Terraform AWS Provider: aws_db_instance (argument
engine_version) in v4.0.0+. - Amazon ElastiCache Serverless for Valkey: ElastiCache Serverless for Valkey now offers powerful vector, full-text, and hybrid search capabilities, along with aggregations. This enables applications to achieve microsecond latency and high recall for vector search, supporting billions of queries daily. Updates are searchable almost immediately after writes. This feature significantly enhances ElastiCache's utility for AI/ML workloads, particularly for real-time RAG architectures. Security considerations include ensuring robust network isolation (VPC), encryption at rest (KMS), and in transit, alongside granular IAM policies for access control. *Terraform Support*: Supported in Terraform AWS Provider: aws_elasticache_serverless_cache (inherent capability of the Valkey engine version) in v5.82.0+.
Storage & Data Lake Enhancements
- Amazon S3 Tables: Amazon S3 Tables now integrate with S3 server access logging, providing detailed audit trails for requests to all tables and namespaces. Logs capture essential information such as requester, operation, table name, response status, and timing, using a familiar space-delimited format. This is a critical enhancement for security and compliance, enabling organizations to reuse existing log analysis workflows with tools like Amazon Athena and DuckDB for comprehensive auditing and incident response. *Terraform Support*: Supported in Terraform AWS Provider: aws_s3tables_table_bucket (block
logging_configuration) in v5.80.0+. - Amazon S3 Access Grants: The
ListCallerAccessGrantsAPI now supports anInstanceScopeparameter set toALL, allowing discovery and listing of all grants across multiple AWS accounts, including those shared via AWS RAM. TheGetDataAccessAPI automatically routes requests to the owning account for matching grants. This significantly improves visibility and auditability of S3 access permissions in complex multi-account environments, streamlining access reviews and enhancing overall security posture by making it easier to identify and manage access. *Terraform Support*: Pending provider release for specific API parameter exposure (use AWS CLI/SDK or custom provisioner in the interim). Theaws_s3_access_grants_instanceresource manages the instance, but this specific API enhancement for listing across accounts isn't a direct resource attribute. Terraform: aws_s3_access_grants_instance.
AI/ML & Analytics Innovations
- Amazon Bedrock AgentCore: The
InvokeWebSearchWeb Search API is now generally available on Amazon Bedrock AgentCore. This API enables a single-call web search, returning ranked results with text extracts, page titles, URLs, and publication times. It supports filtering by domain and publication time, empowering AI agents with real-time external information. Security considerations include careful IAM policy management forInvokeWebSearchto control access, VPC PrivateLink for secure, private connectivity, and CloudTrail data events for comprehensive auditing of web search invocations. *Terraform Support*: Pending provider release for specific API parameter exposure (use AWS CLI/SDK or custom provisioner in the interim). Theaws_bedrockagent_agentresource manages the agent itself, but this specific API enhancement isn't a direct resource attribute. Terraform: aws_bedrockagent_agent. - Amazon OpenSearch Service: OpenSearch Service now offers clone-based upgrades, enabling validation and rollback of version upgrades. This feature creates a cloned cluster with data copied from the primary, allowing the primary to continue serving traffic while the new version is validated. After successful validation, traffic can be cut over to the clone, with the option to retain the old version for a chosen period to facilitate rollbacks. This significantly reduces the risk associated with major version upgrades, enhancing operational stability, security, and data integrity. *Terraform Support*: Supported in Terraform AWS Provider: aws_opensearch_domain (management of
engine_versionand implicit upgrade process) in v5.82.0+.
Management & Governance Updates
- Amazon CloudWatch: CloudWatch PromQL alarms can now directly trigger Amazon EC2 and Application Auto Scaling step and simple scaling policies. This allows organizations to use consistent PromQL queries for both monitoring and dynamic resource scaling, enhancing operational automation and resilience. Security implications include improved availability and performance, which can help mitigate certain types of DoS attacks by automatically scaling resources. Proper IAM roles and permissions for CloudWatch and Auto Scaling are crucial for secure operation. *Terraform Support*: Supported in Terraform AWS Provider: aws_cloudwatch_metric_alarm (arguments like
alarm_actionsandmetric_query) in v5.49.0+. - Amazon CloudWatch Logs: CloudWatch Logs now supports account-level retention policies, allowing a single policy to define retention for all or prefix-matched log groups (e.g.,
/aws/lambda). When multiple policies match a log group, the longest retention period is applied. This feature simplifies compliance with data retention regulations (e.g., GDPR, HIPAA, PCI DSS) across an entire AWS account, ensuring consistent application of rules and reducing the risk of accidental data loss or non-compliance. Centralized management also improves auditability and reduces configuration drift. *Terraform Support*: Supported in Terraform AWS Provider: aws_cloudwatch_log_account_policy (argumentspolicy_name,policy_type,policy_document) in v5.67.0+.