All news in category “Vendor and Hyperscaler Watch”

🔧 Amazon Elastic Container Service now automatically detects and repairs container instances with impaired agent connectivity across all compute options. It surfaces a new AGENT_CONNECTIVITY health change event for AWS Fargate, Amazon ECS Managed Instances, and Amazon ECS on EC2. For ECS Managed Instances and Fargate, ECS will automatically drain tasks and launch replacement capacity while deregistering impaired instances; ECS on EC2 users can use the event to trigger replacements. This capability is available at no additional cost in all AWS Commercial and GovCloud (US) Regions.

🔒 AWS Glue now supports catalog federation for remote Iceberg REST catalogs in AWS GovCloud (US) Regions, enabling direct, secure access to Iceberg tables stored in Amazon S3 and cataloged remotely. With this feature, analytics engines can query remote Iceberg tables without copying data, and metadata synchronizes in real time between AWS Glue Data Catalog and the remote catalog. The capability integrates with AWS Lake Formation for fine-grained access controls and cross-account sharing, and is supported by Amazon Redshift, Amazon EMR, Amazon Athena, AWS Glue, and third-party engines. Catalog federation is available via the Lake Formation console and the AWS Glue and Lake Formation SDKs/APIs in GovCloud (US-East) and GovCloud (US-West).

🔐 SageMaker MLflow now supports customer-managed keys (CMK) via AWS Key Management Service (KMS). This enhancement lets organizations with strict security or compliance needs manage encryption keys themselves and gain enhanced control and auditing through AWS CloudTrail. Customer-managed keys must be symmetric and created in the same AWS account and region as the MLflow App. The feature is generally available in all Regions where MLflow App is offered.

📝 Cloudflare migrated its blog to EmDash, a CMS built for Astro and Cloudflare, moving on August 12. The redesign added dark mode, Kumo-aligned frontend patterns, and improved caching and performance. A staged rollout with a proxy Worker ensured zero downtime while enabling new agent-friendly features like a Model Context Protocol (MCP) server.

🔒 Amazon Elastic Kubernetes Service (Amazon EKS) now lets you associate up to 10 external OpenID Connect (OIDC) identity providers with a single cluster. This enables distinct user populations—employees, contractors, CI/CD systems—to authenticate directly without consolidating providers or using an identity broker. Each provider is configured and managed independently and coexists with existing IAM authentication. You add providers via the AWS Management Console or the AssociateIdentityProviderConfig API at no extra cost in all EKS regions.

🔒 The Trusted Computing Group (TCG) published guidance on August 24 to help organizations verify that trusted platform modules (TPMs) meet post-quantum cryptography (PQC) requirements. The guidance complements the TCG PC Client Platform TPM Profile (PTP) 1.07, developed with contributions from major vendors, and defines two readiness categories: TCG PQC-ready TPM and TCG PQC-upgradable TPM. TCG also plans to extend its certification programs to cover PQC readiness.

🔔 Amazon Aurora PostgreSQL-Compatible Edition now supports PostgreSQL 18.4, 17.10, 16.14, 15.18, and 14.23 bringing community bug fixes and Aurora-specific improvements. Automatic minor version upgrades can be applied during scheduled maintenance windows and can be orchestrated across accounts using the AWS Organizations Upgrade Rollout Policy. Enable automatic upgrades to address known CVEs and simplify large-scale operations. Aurora continues to offer high performance, global resilience, serverless scale-to-zero compute, and improved I/O price-performance.

🛠️ Amazon SageMaker HyperPod now integrates Ray with built-in observability, resilient distributed training, accelerated inference, and managed development environments. Data scientists can create and manage Ray clusters from SageMaker Studio, attach JupyterLab or a local IDE for interactive iteration, and use Grafana and Amazon Managed Service for Prometheus for one-click observability. HyperPod provides node auto-recovery, hung job detection, tiered checkpointing, and task governance to improve GPU utilization and reliability, plus a tiered KV cache and JumpStart model deployment for faster Ray Serve inference.

🔐 Amazon MSK Replicator now supports OAuth 2.0 (SASL/OAUTHBEARER) authentication for replicating data from external Apache Kafka clusters—including on-premises, self-managed on AWS, or other clouds—to Amazon MSK Provisioned clusters. This enables replication from clusters using OAuth/OIDC for migration, disaster recovery, and hybrid or multi-cloud data distribution. MSK Replicator continues to automate replication without custom tooling and preserves topic names while preventing replication loops and synchronizing consumer offsets.

🔍 Amazon Connect Customer now supports automated information extraction from voice and chat interactions, capturing verbatim data like account numbers and derived insights such as reason for contact and next steps. Administrators define conversational analytics rules and extraction runs on raw content before redaction, enabling capture of sensitive items while still supporting redaction in recordings and transcripts. Extracted values appear to agents in After Contact Work, are searchable by supervisors, and are available to developers via APIs, Kinesis Data Streams, and S3 outputs.

🛡️ Microsoft is introducing a Teams meeting protection policy that lets administrators automatically block identified external bots from joining meetings. This builds on a June update that flagged detected bots for organizer approval; the new policy prevents such bots from joining without confirmation. Rolling out in targeted release through August and GA by late September, the setting is off by default and must be enabled and assigned via the Teams admin center.

🛠️ AWS ParallelCluster 3.16 is now generally available and introduces pcluster-diag, an on-node diagnostics tool included in ParallelCluster AMIs that produces structured reports to simplify issue identification. The release also improves cluster lifecycle resilience with more robust creation, updates, and image builds. The HPC and AI/ML software stack receives updates to NVIDIA drivers, CUDA, EFA installer, and Slurm.

🤝 Amazon Bedrock now offers OpenAI's GPT-5.6 family — Terra and Luna — in AWS GovCloud (US-West) and (US-East), delivered via Bedrock's next-generation inference engine for security and high performance. Terra balances capability and cost while Luna prioritizes fast, low-cost inference; both support 1 million token context windows and prompt caching with explicit breakpoints for reduced billing on repeated context. Use the Bedrock Console or Responses API on the bedrock-mantle endpoint to get started.

🛠️ Microsoft confirmed that August 2026 .NET Framework cumulative updates are causing some Windows Presentation Foundation (WPF) applications to fail with a System.IO.FileFormatException when printing or generating PDF/XPS content using certain fonts such as Calibri. The issue affects multiple client and server Windows releases, and Microsoft provided a temporary AppContext switch workaround that must be added to the application config file. The company warns this workaround disables protections added in the update and should be used only temporarily while a permanent fix is developed.

🎮 Microsoft provided a temporary workaround for a gaming crash issue introduced with the August 11, 2026 Windows 11 updates (KB5121003). The bug caused crashes, freezes, EXCEPTION_ACCESS_VIOLATION errors, and restarts when launching certain games on 24H2 and 25H2 systems. Investigation linked the problem to RGB device drivers/components, specifically files like inpoutx64.sys. Microsoft’s interim fix disables the driver via a Registry change, with warnings to back up and potential loss of RGB functionality.

🚀 Starting today, Amazon RDS for MySQL supports community MySQL minor version 8.4.11, which includes operational improvements and introduces post-quantum TLS (PQ-TLS) key exchange for enhanced in-transit encryption options. AWS recommends upgrading to newer minor versions to receive CVE fixes, performance improvements, and bug fixes provided by the MySQL community. You can use automatic minor version upgrades during maintenance windows or Amazon RDS Managed Blue/Green deployments for safer updates.

🛡️ Cloudflare introduces Bot Preference Sync to align zone-level AI bot settings with a site's robots.txt, available to all customers from Free to Enterprise. The feature prepends generated directives to any existing robots.txt so owners' Search, Agent, and Training preferences match edge-enforced rules and managed blocks. Bot Preference Sync is on by default for new customers and can be disabled by those who need custom, fine-grained policies.

🚀 AWS Neuron 2.32.0 delivers NKI 0.6.0 with on-device top-K, variable-length all-gather, and new loop constructs for data-dependent iteration. The release includes 13 new NKI Library kernels for Mixture of Experts (MoE) training and sparse attention, plus PyTorch references for 22 additional kernels. Runtime collectives now support variable-size all-gather, reduce-scatter, and all-to-all on Trn2 and Trn3, and the vLLM Neuron plugin moves to vLLM 0.24.0 across Neuron AMIs and containers.

🔔 Amazon Bedrock announces reduced pricing for OpenAI GPT-5.6 Sol, reflecting OpenAI's API price cuts. Sol now costs $4 per million input tokens and $20 per million output tokens, representing a 20% reduction on input and a 33.3% reduction on output pricing. The promotional rates apply at least through November 21, 2026, making high-volume, agentic coding and research workflows more cost-effective.

🤖 Managers can now query Amazon Connect Customer in plain language to get answers, evidence, and recommended fixes in seconds. The feature searches over 150 metrics across self-service, agent, and queue performance to identify what matters and explain why. Users can start with broad questions and drill down—receiving prioritized actions with confidence scores and projected impact. Available in all Regions that support Amazon Connect Customer AI Agents.