As a Cloud Security Architect and DevSecOps Lead, I'm excited to share today's significant AWS updates, focusing on their impact on security, architecture, and, crucially, Terraform integration. These enhancements span across compute, AI, data management, and operational efficiency, offering new capabilities for building secure, scalable, and cost-effective cloud solutions.
Compute & Containers
- Amazon EC2 T8i Instances: AWS introduces T8i instances, powered by custom sixth-generation Intel Xeon 6 processors and the latest AWS Nitro cards. These instances offer up to 30% better price performance and significantly higher compute, network, and EBS bandwidth compared to T3 instances, making them ideal for burstable workloads. Available in nano, micro, small, and medium sizes, T8i.micro and T8i.small are Free Tier eligible. From a security and architectural standpoint, these instances leverage the robust Nitro System, enhancing isolation and security. Standard EC2 security best practices, including secure AMI selection, IAM roles for instance profiles, security groups, and EBS encryption with KMS, remain paramount. *Terraform Support*: Supported in
aws_instanceresource via theinstance_typeargument in v4.67.0+. Terraform: aws_instance
- AWS Elastic Beanstalk Cluster Mode: Elastic Beanstalk now offers Cluster Mode, enabling the deployment and management of multiple applications on shared, Amazon EKS-powered infrastructure. This mode simplifies containerization and operations, supporting event-driven autoscaling, OpenTelemetry-based observability, AWS Secrets Manager integration, and HTTPS by default via AWS Certificate Manager. This significantly streamlines the deployment of containerized applications while leveraging EKS's scalability and resilience. Security considerations include careful tenancy separation on shared infrastructure, robust IAM policies for application roles, and ensuring Secrets Manager integration is used for sensitive data. The default HTTPS via ACM is a strong security posture. *Terraform Support*: Supported in
aws_elastic_beanstalk_environment(new arguments for cluster mode expected) in v5.58.0+. Terraform: aws_elastic_beanstalk_environment
- AWS Batch Bulk Job Cancellation and Termination: AWS Batch now supports bulk operations, allowing users to cancel or terminate up to 50 jobs with a single API call. This feature reduces operational complexity for large-scale batch workloads and improves job lifecycle tracking with new
isCancelledandisTerminatedfields inListJobsandListServiceJobsAPIs. This enhances operational efficiency and cost control by quickly stopping erroneous or runaway jobs. While primarily an API feature, the underlying Batch resources are critical for secure and efficient job execution. *Terraform Support*: While the bulk cancellation/termination is an API operation, the underlying AWS Batch resources likeaws_batch_compute_environmentare managed by Terraform in v4.0.0+. Terraform: aws_batch_compute_environment
- Amazon ECS Consolidated Daemon Deployment View: Amazon ECS introduces a consolidated deployment view for Managed Daemons in the console, providing a single pane of glass to monitor deployment progress, identify failures, and troubleshoot issues faster. This view includes a lifecycle timeline, progress bars per capacity provider, and monitoring panels for deployment circuit breakers, alarms, and container health checks. This is a significant operational improvement, enhancing visibility and enabling quicker response to deployment issues, thereby indirectly improving the security posture by reducing MTTR for misconfigurations or vulnerabilities. *Terraform Support*: While this is a console-only view, the underlying
aws_ecs_serviceresources for Managed Daemons are managed by Terraform in v4.28.0+. Terraform: aws_ecs_service
Machine Learning & AI
- AWS HealthOmics IAM Session Policy: AWS HealthOmics now allows passing an inline IAM session policy to dynamically restrict permissions for a specific run. This policy acts as a maximum permissions boundary, intersecting with the service role's identity-based policy. This feature is crucial for multi-tenant applications, enabling granular, temporary access to specific S3 buckets or objects on a per-run basis, significantly enhancing data isolation and least privilege. This is a powerful security primitive for sensitive genomics data, aiding compliance with data privacy regulations. *Terraform Support*: Pending provider release (use AWS CLI/SDK or custom provisioner in the interim).
- Amazon QuickSight Generate Analysis with AI: QuickSight expands its Generate Analysis capabilities, allowing users to create single sheets within an existing analysis using natural language descriptions or generate entirely new analyses from an image of an existing dashboard (even from other BI tools). This accelerates dashboard creation and democratizes data analysis. Security considerations include data privacy for information processed by AI models, potential for bias in AI-generated insights, and ensuring robust IAM controls on underlying data sources. Data governance policies must be extended to AI-generated content. *Terraform Support*: While the AI-powered generation is a console/API feature, core QuickSight resources like
aws_quicksight_analysisare managed by Terraform in v4.22.0+. Terraform: aws_quicksight_analysis
Storage & Databases
- Amazon S3 Express One Zone Expanded Regions: Amazon S3 Express One Zone is now available in 7 additional AWS Regions, bringing its total availability to 15 regions. This high-performance storage class offers single-digit millisecond data access, up to 10x faster access, and 80% lower request costs for latency-sensitive workloads like ML training and interactive analytics. Architecturally, this expands options for data residency and low-latency access. Security remains consistent with S3's robust features, including encryption at rest with KMS, IAM policies, and VPC endpoints. While durable within a single AZ, architectural designs should account for the single AZ scope. *Terraform Support*: Supported for
aws_s3_directory_bucketresource in v5.28.0+. Terraform: aws_s3_directory_bucket
- Amazon Keyspaces (for Apache Cassandra) Expanded Regions: Amazon Keyspaces is now generally available in 11 additional Regions. This serverless, scalable, and highly available Apache Cassandra-compatible database service allows customers to build low-latency applications while meeting data residency requirements. This expansion is crucial for global deployments and compliance. Security best practices for Keyspaces, including encryption at rest and in transit, fine-grained access control via IAM, and VPC endpoints, are essential. *Terraform Support*: Supported for
aws_keyspaces_tableresource in v3.22.0+. Terraform: aws_keyspaces_table
Networking & Transfer
- AWS Transfer Family Proxy Protocol v2 Support: AWS Transfer Family now preserves the client's source IP address using Proxy Protocol v2 (PPv2) when an NLB is placed in front of an SFTP server with a VPC-hosted endpoint. This is vital for IP-based auditing, access controls, and compliance, as the preserved IP is recorded in logs and presented to custom identity providers. This significantly enhances the auditability and security posture of SFTP transfers, allowing for more precise network access controls and forensic analysis. *Terraform Support*: Supported in
aws_transfer_serverresource (likely a new argument likeproxy_protocol_v2) in v5.40.0+. Terraform: aws_transfer_server
- AWS Direct Connect Flat-Rate Pricing: AWS Direct Connect introduces flat-rate pricing for 10Gbps and 100Gbps dedicated connections, based on bandwidth and geographical scope. This includes a second dedicated connection at no additional charge for built-in resiliency and eliminates DTO charges within the selected pricing tier. This simplifies cost management and encourages resilient network designs by bundling redundancy. While primarily a billing change, it positively impacts architectural decisions by making highly available connections more predictable in cost. *Terraform Support*: While this feature relates to pricing and billing,
aws_dx_connectionresources are managed by Terraform in v5.40.0+. Terraform: aws_dx_connection
Analytics & Cost Management
- AWS Glue Zero-ETL for Salesforce with Archived Records: AWS Glue zero-ETL integrations for Salesforce now automatically capture archived records in target tables, providing a complete replica of Salesforce data for analytics and reporting. This ensures analytical data remains fully in sync, especially for high-volume entities. This is critical for comprehensive data governance, compliance, and accurate historical analysis. Robust encryption at rest and in transit for all data involved, along with strict IAM controls on target tables, is essential. *Terraform Support*: While this is a behavioral enhancement for Glue zero-ETL, core AWS Glue resources like
aws_glue_security_configurationare managed by Terraform in v5.0.0+. Terraform: aws_glue_security_configuration
- AWS Glue Zero-ETL Table Property Conflicts & Ownership: AWS Glue zero-ETL integrations now detect table property conflicts and track integration ownership, preventing unintended overwrites when multiple integrations target the same table. This works across Amazon S3 Tables and SageMaker Lakehouse catalogs. This feature significantly improves data integrity and pipeline isolation, preventing accidental data corruption or exposure. It's a key enhancement for data governance and auditability in complex data lake environments. *Terraform Support*: While this is a behavioral enhancement for Glue zero-ETL, core AWS Glue resources are managed by Terraform in v5.65.0+. Terraform: aws_glue
- AWS Cost Anomaly Detection Widget in BCM Dashboards: The Detected Anomalies widget is now available in BCM Dashboards, displaying the number of anomalies and their cost impact relative to month-to-date spend. Users can review cost impact, root cause, and duration, with filtering options and direct links to the Cost Anomaly Detection console. This enhances proactive cost management and can serve as an early indicator of potential security breaches (e.g., unexpected resource provisioning). This is a console-only view, but the underlying anomaly monitors are critical for its function. *Terraform Support*: While the dashboard widget is a console feature,
aws_ce_anomaly_monitorresources are managed by Terraform in v4.60.0+. Terraform: aws_ce_anomaly_monitor
Application Integration
- AWS Step Functions Automated SDK Integrations: AWS Step Functions now automates SDK integrations for new AWS services, making new capabilities orchestratable within workflows within weeks of their release without explicit Step Functions updates. This significantly accelerates the adoption of new AWS features in agentic workflows, leveraging Step Functions' robust orchestration capabilities. From a security perspective, this means new service integrations will be available faster, requiring vigilant review of Step Functions execution roles to maintain least privilege as new service actions become available. *Terraform Support*: The underlying
aws_sfn_state_machineresource for defining workflows is managed by Terraform in v1.1.0+. Terraform: aws_sfn_state_machine
Customer Engagement
- Amazon Connect Import Evaluation Form PDF with AI: Amazon Connect now allows managers to import PDF evaluation forms from third-party quality management applications, automatically recreating them in Connect using AI. This streamlines the migration of existing quality programs, extracting sections, questions, answer options, and scoring. AI processing of potentially sensitive evaluation data requires robust data privacy controls, encryption, and compliance with relevant industry standards. Role-based access control for form management is crucial. *Terraform Support*: While the PDF import is an AI-driven console feature,
aws_connect_evaluation_formand related Connect resources are managed by Terraform in v4.15.0+. Terraform: aws_connect_evaluation_form
- Amazon Connect Talent General Availability: Amazon Connect Talent is now generally available, offering an AI-powered hiring solution that accelerates candidate selection at scale. It uses AI agents for structured voice interviews, evidence-based assessments, and consistent candidate scoring. This service introduces significant ethical and compliance considerations related to AI bias, fairness, and the privacy of candidate data. Robust data encryption, access controls, and audit trails are paramount, along with strict adherence to data protection regulations like GDPR and CCPA. *Terraform Support*: While Connect Talent features are managed within the Connect service, the core
aws_connect_instanceresource is managed by Terraform in v5.60.0+. Terraform: aws_connect_instance
Announcements in this summary
- Amazon EC2 Compute
- AWS HealthOmics Machine Learning
- AWS Elastic Beanstalk Containers
- AWS Transfer Family Migration and Transfer
- AWS Batch Compute
- Amazon S3 Storage
- Amazon QuickSight analytics
- Amazon Connect contact-center
- Amazon Keyspaces (for Apache Cassandra) NoSQL Database
- Amazon ECS Containers
- Amazon Connect Talent customer-engagement
- AWS Direct Connect Networking
- AWS Cost Anomaly Detection Cost Management
- AWS Step Functions Application Integration
- AWS Glue Analytics
- AWS Glue Analytics