Cloud · Top stories
Google Cloud introduces Fault Injection Testing in public preview
Google Cloud has launched Fault Injection Testing in public preview, allowing users to automate failure testing in their cloud environments. This tool helps ensure applications remain available and reliable during disruptions by deliberately introducing faults to verify safety mechanisms.
AWS Introduces Specification-Driven Composition for Flexible Data Workflows
AWS has introduced a specification-driven composition pattern for building data transformation workflows, which separates workflow intent from processing logic. This approach aims to reduce duplicated pipeline code and simplify validation and governance in data pipelines, particularly for organizations managing numerous datasets and workflow variations.
AWS Acquires DuckLabs, DuckDB Open-Source Projects to Remain Independent
Amazon Web Services (AWS) is acquiring DuckLabs, the company behind the open-source analytical database DuckDB. The acquisition, effective in early September, will integrate the DuckLabs team into AWS while ensuring DuckDB and related projects remain open source under the MIT license, with continued stewardship by the nonprofit DuckDB Foundation.
AWS Releases Independent Report on Landing Zone Accelerator for Australian Government ISM Compliance
AWS has released an independent assessment report on how Landing Zone Accelerator (LZA) on AWS can deploy multi-account environments with Australian Government Information Security Manual (ISM) security controls. This report, conducted by gwi.digital, provides Australian customers with a documented foundation to accelerate IRAP assessment readiness, which is crucial for public sector, defense, and critical infrastructure agencies.
Amazon EC2 Celebrates 20th Anniversary, Highlighting Evolution and Expansion
Amazon EC2 marks its 20th anniversary, reflecting on its growth from a single instance type in one region to over 1,200 instance types across 39 regions globally. This milestone highlights EC2's role in shaping cloud computing infrastructure and its continued expansion into new form factors and specialized hardware for AI workloads.
Cloudflare Launches Bot Preference Sync to Automate robots.txt Updates for AI Bot Management
Cloudflare introduced Bot Preference Sync, a new feature that automatically updates a website's robots.txt file to reflect AI bot configurations set within Cloudflare's platform. This aims to simplify managing how different AI bot categories, such as search, agent, and training crawlers, interact with website content, ensuring consistency between stated preferences and enforcement rules.
AWS Glue 6.0 Released with 30% Price Reduction and Full Apache Iceberg v3 Support
AWS has released Glue 6.0, which offers a 30% price reduction compared to previous versions and includes full support for Apache Iceberg v3 features. This update provides a modernized runtime with Apache Spark 4.1, Python 3.12, and Scala 2.13, aiming to improve performance and simplify ETL authoring.
AWS Marketplace Sees Significant Growth in AI Agent Offerings and Usage
AWS Marketplace reports a substantial increase in AI agent offerings, growing from 1,000 to over 4,000 in one year, and AI agents are now the third most searched keyword. This growth reflects a broader trend of AI agents automating administrative and due diligence tasks within software acquisition and deployment processes.
Codex on AWS Bedrock Lacks Explicit Cache Controls, Causing High GPT-5.6 Sol Costs
A bug in the native Amazon Bedrock provider for Codex CLI prevents explicit prompt caching for GPT-5.6 Sol, leading to significantly higher costs due to excessive cache-write tokens. This issue impacts agentic coding workloads where stable instructions are followed by changing content, which should benefit from caching.
AWS Network Firewall adds rule hit count support for stateful rules
AWS Network Firewall now includes a rule hit count feature that tracks how often stateful rules match network traffic. This capability helps security teams identify unused rules, improve incident response, and validate security control effectiveness for compliance frameworks like PCI 4.0 and DORA.
Multiple Apple Services Including Find My and iCloud Experiencing Outages for Some Users
Apple's System Status page confirms that several services, including Find My, Game Center, Photos, and various iCloud functions, are currently experiencing issues for some or all users. This outage impacts core functionalities for Apple device users and developers, affecting accessibility and productivity across the ecosystem.
Microsoft 365 search outage affects Outlook, SharePoint Online, and OneDrive users
Microsoft confirmed an outage affecting search functionality in Microsoft 365 applications, including Outlook, SharePoint Online, and OneDrive. The issue stemmed from a recent deployment that caused resource utilization problems, impacting some users attempting to search for content.
PJM Proposes Rules to Prioritize Residential Power Over New Large Data Centers During Shortages
PJM Interconnection, the grid operator for 13 states and Washington D.C., has asked federal regulators to approve new rules that would allow it to cut power to new data centers (50 MW or more) before residential customers during electricity shortages. This proposal aims to address insufficient generation capacity following failed capacity auctions and projected load growth, requiring new large data centers to secure their own power supply to avoid curtailment.
Cherokee Nation Bans Hyperscale Data Center Development on Tribal Lands
The Cherokee Nation has prohibited hyperscale data center development on its tribally owned and trust lands following a task force report indicating citizen opposition due to concerns about energy, water, and environmental impact. This decision reflects a growing trend of local jurisdictions and tribal nations imposing moratoriums or restrictions on data center construction.
Cloudflare Launches Cache Response Rules for Post-Origin Cache Control
Cloudflare introduced Cache Response Rules, a new engine that processes origin server responses before content is cached, allowing for modifications to cache-related headers and improved cache hit rates. This feature provides finer control over caching behavior without requiring changes to the origin application, which can simplify CDN migrations and reduce infrastructure costs.
Cloudflare Migrates cdnjs to Its Developer Platform, Serving 9 Billion Daily Requests
Cloudflare has fully migrated its cdnjs open-source CDN for JavaScript and CSS libraries to its Developer Platform, replacing a distributed infrastructure that included Google Cloud Platform services. This migration consolidates cdnjs onto Cloudflare's own serverless products, demonstrating the platform's scalability and capabilities for a service handling 9 billion daily requests.
Namecheap experiences outage due to data center cooling failure in Phoenix
Hosting provider Namecheap suffered an outage affecting multiple services after a cooling system failure at its Phoenix data center. The incident, caused by storm damage, impacted hosting plans, DNS, and email, with services expected to begin returning by 3:00-3:30 PM ET. This disruption highlights the vulnerability of cloud services to physical infrastructure failures and environmental events.
AWS introduces IAM role manager to automate IAM role creation for services
AWS has launched IAM role manager, a new feature that automates the creation and configuration of IAM roles when building applications in supported service consoles. This feature simplifies the initial setup process for developers by automatically provisioning necessary roles, allowing them to refine permissions later.
Netflix Replaces In-House Batch Job System with Open-Source Kueue
Netflix has migrated most of its batch workloads from its proprietary Compute Managed Batch (CMB) system to Kueue, an open-source cloud-native job queueing system. This shift was driven by Kueue's advanced capabilities and better integration with the Kubernetes ecosystem, which CMB lacked as it aged.
Spotify Introduces Random Access Parquet (RAP) for Low-Latency Queries on Data Lake
Spotify developed Random Access Parquet (RAP), a new storage architecture that enables low-latency point queries directly on data lake files without replicating datasets into operational databases. This allows online services and AI applications to retrieve individual records efficiently from existing analytical datasets, reducing storage costs and data duplication.
Amazon SageMaker HyperPod introduces tiered KV cache with Curvine for LLM inference
Amazon SageMaker HyperPod now supports a tiered KV cache architecture using Curvine, a distributed cache filesystem, to extend KV cache beyond GPU and CPU memory into a shared NVMe pool. This development allows for KV cache reuse across replicas, improving time-to-first-token (TTFT) and reducing infrastructure costs for large language model (LLM) inference.
Google Cloud DMS Automates SQL Server to PostgreSQL Migration with Gemini AI
Google Cloud's Database Migration Service (DMS) now automates the translation of SQL Server stored procedures that return multiple result sets into PostgreSQL's `SETOF refcursor` architecture. This update integrates Gemini, Google's generative AI, to convert proprietary SQL code to PostgreSQL PL/pgSQL, addressing a key bottleneck in large-scale database migrations.
Google expands Americas network with three new subsea cables and Firmina branch
Google is expanding its network infrastructure in the Americas with three new subsea cable systems: Alisios, Canoa, and OlaLuz, along with a new branch for the existing Firmina cable. This expansion aims to improve connectivity, reliability, and resilience for internet users and cloud services across Latin America, the Caribbean, and North America.
Google Cloud launches Developer Device Platform for mobile app development and testing
Google Cloud has launched the public preview of its Developer Device Platform (DDP), a fully managed cloud service providing on-demand access to physical devices and virtual emulators for mobile app development and testing. This platform aims to simplify the process of building, testing, and debugging applications across various devices, particularly for agentic development.
Amazon EKS Reduces Multi-Gigabyte Container Image Pull Times to Seconds
Amazon EKS has reduced the time it takes to pull multi-gigabyte container images from several minutes to seconds, particularly benefiting machine learning workloads. This improvement was achieved by optimizing the image pull pipeline to better utilize existing network bandwidth, storage throughput, and compute resources on accelerated instances.
Google Kubernetes Engine Integrates ClusterNetworkPolicy for Centralized Network Security
Google Kubernetes Engine (GKE) has integrated ClusterNetworkPolicy (CNP), an open-source standard from the Kubernetes SIG-Policy Working Group, to provide centralized network security management. This integration allows administrators to implement consistent, non-bypassable policies across a cluster, addressing the limitations of standard Kubernetes NetworkPolicy for global security enforcement.
Canva Redesigns Session Revocation Using Amazon S3 for Scalability
Canva re-engineered its session revocation infrastructure to handle hundreds of millions of active sessions by storing revocation data in Amazon S3 and distributing it as in-memory indexes to application gateways. This change improved deployment speed, reduced database infrastructure, and decreased the memory footprint of its revocation cache by 87.5%. The new architecture allows Canva to scale its authentication system more efficiently without relying on constant networked database lookups.
GeForce NOW exploit grants full Windows desktop access, enabling local AI model execution
A modder discovered an exploit in Nvidia's GeForce NOW Ultimate tier that allows users to access the full Windows desktop environment, bypassing the service's gaming-only restrictions. This exploit enables the execution of local AI models and other desktop applications, transforming the cloud gaming service into a general-purpose cloud PC. The method involves manipulating game files and is against Nvidia's terms of service, risking account bans.
Pinterest Introduces Resource Provisioner Pipeline for Secure AWS Infrastructure Management
Pinterest developed and released the Resource Provisioner Pipeline (RPP), a custom Terraform execution engine, to manage its AWS infrastructure securely. RPP enforces least-privilege access and dual-control reviews across thousands of cloud resources, addressing security challenges in a multi-repository environment. This system enhances security and consistency for Pinterest's AWS operations by centralizing Terraform execution and applying strict guardrails to GitHub Actions workflows.
Snowflake Introduces Data Mirroring for Postgres Replication into Snowflake
Snowflake has launched Data Mirroring, a new feature in public preview for its Postgres service, which reimagines Postgres replication to push changes directly into Apache Iceberg tables and then into Snowflake. This feature aims to provide highly resilient data replication with low cost, low lag, and transactional consistency, simplifying the process of moving data from transactional databases to analytical databases.
Amazon Data Center Construction in Gilroy, CA, Began Under 45-Year-Old Zoning Rules
Amazon began construction on a data center in Gilroy, California, utilizing local zoning rules from 45 years ago, which allowed the project to proceed without recent public comment. This situation highlights a growing tension between community concerns, particularly regarding water usage, and the rapid expansion of data center infrastructure.
BigQuery Data Transfer Service Adds New Connectors and Capabilities for Data Ingestion
Google Cloud's BigQuery Data Transfer Service (DTS) has expanded its data ingestion capabilities with new connectors and features, including direct ingestion into Apache Iceberg tables, a managed Model Context Protocol (MCP) Server, and support for additional databases and marketing platforms. These updates aim to reduce the effort required for data movement into BigQuery, allowing users to integrate more diverse data sources with less manual setup.
AWS Restricts Internal EC2 Use Due to High CPU Demand from AI Workloads
Amazon Web Services is limiting its engineers' access to EC2 instances to conserve CPU capacity for paying customers. This internal crackdown is a direct result of surging CPU demand driven by complex agentic AI workloads, which require more CPU resources than traditional GPU-centric AI inference.
Amazon Cognito Introduces Self-Service Provisioned Limits for Rate Limit Management
Amazon Cognito now offers provisioned limits, allowing users to adjust authentication rate limits on-demand through the AWS Management Console. This change reduces the time required for capacity increases from up to two weeks to minutes, enabling better cost optimization and responsiveness to traffic fluctuations.
Oracle Halves Always Free ARM Limits for Cloud Infrastructure
Oracle is reducing the "Always Free" Ampere A1 ARM instance limits on its Oracle Cloud Infrastructure (OCI) from 4 OCPUs and 24 GB RAM to 2 OCPUs and 12 GB RAM, effective August 18, 2026. Users exceeding these new limits must adjust their instances or face automatic termination, impacting those who relied on the previous, more generous free tier for development and testing.
Google Cloud Filestore now uses Colossus for enhanced scalability and performance decoupling
Google Cloud's Filestore, an NFS file service, now incorporates Colossus, Google's distributed storage system, as its backend storage layer. This integration allows Filestore to decouple storage capacity and performance, offering greater flexibility and scalability for demanding workloads, including AI and containerized environments.
YugabyteDB Launches AMP for Serverless PostgreSQL to Manage AI Agent Database Sprawl
YugabyteDB introduced Agentic Multitenant Postgres (AMP), a new serverless PostgreSQL tier designed to manage the proliferation of databases created by AI agents. This solution addresses the scaling challenge of numerous small, isolated databases rather than a single large one, offering cost efficiency by scaling to zero when idle.
AWS and Superblocks Partner to Embed Vibe-Coding Tool in AWS Private Clouds
Superblocks, a vibe-coding startup, announced a multi-year joint marketing agreement with Amazon Web Services (AWS) to embed its tool within AWS customer private clouds. This partnership allows enterprise users to utilize Superblocks for vibe coding with data remaining within their AWS environment, integrating with services like Amazon Aurora and Amazon Bedrock, and falling under IT management. The collaboration signifies a broader trend where hyperscalers aim to keep enterprise AI scaffolding within their cloud ecosystems.
Google Data Commons on Spanner Graph Reaches General Availability, Platform Previewed
Google announced the general availability of Data Commons on Spanner Graph and a preview of the new Data Commons Platform. This update allows businesses to integrate their private data with public datasets from over 100 providers, facilitating the creation of knowledge graphs for informed decision-making.
Amazon Reaches $3 Trillion Market Cap Following Strong Q2 Earnings and Cloud Growth
Amazon's market capitalization exceeded $3 trillion for the first time after its stock rose following a better-than-expected Q2 earnings report. This surge was driven by strong growth in Amazon Web Services (AWS) revenue, fueled by increased demand for artificial intelligence infrastructure.