Weekly Cloud Info #W31 - 2025

Author

Walid LARABI
August 01, 2025

Hi!

Discover Amazon's new DocumentDB Serverless for cost savings, AWS Lambda's enhanced response streaming, DynamoDB's AI data modeling tool, CloudFront's timeout features, CloudWatch AI observability, Azure's SQL migration tools, and Google Cloud's Vertex AI advancements.

Have a great read. ☕

📰 Top picks of the week

Amazon DocumentDB Serverless Now Generally Available with Automatic Scaling and Cost Savings

Amazon has announced the general availability of Amazon DocumentDB Serverless, a new configuration that automatically scales compute and memory based on application demand, offering significant cost savings. It uses DocumentDB Capacity Units (DCUs) to measure resources and allows for seamless transitions from provisioned instances without data migration. The service supports MongoDB-compatible APIs and is available starting with Amazon DocumentDB version 5.0.0.

AWS Lambda Increases Response Streaming Payload Limit to 200 MB

AWS Lambda response streaming now supports a maximum response payload size of 200 MB, an increase from the previous limit of 20 MB. This enhancement allows for improved performance in latency-sensitive applications by enabling direct processing of larger response payloads without the need for compression or intermediary services. The update benefits use cases such as real-time AI chat and applications handling large datasets or media files. Lambda response streaming is compatible with Node.js managed runtimes and custom runtimes across all supported AWS Regions.

New DynamoDB Data Modeling Tool Accelerates Design Process with AI Integration

Amazon has introduced a new data modeling tool for DynamoDB as part of the DynamoDB Model Context Protocol (MCP) server. This tool leverages AI assistants to streamline the process of creating data models, allowing users to generate models in minutes instead of days. It integrates with various AI platforms, such as Amazon Q Developer and Visual Studio Code, to facilitate efficient data design aligned with application access patterns.

Amazon CloudFront Enhances Origin Timeout Controls with New Features

Amazon CloudFront has introduced two new features: a response completion timeout and custom response timeout values for Amazon S3 origins. These enhancements allow for more precise control over origin response timeouts, improving performance for latency-sensitive applications. Users can now set a maximum wait time for complete responses and customize timeout values beyond the default 30 seconds, applicable across all CloudFront edge locations except AWS China (Beijing).

AWS Launches CloudWatch Generative AI Observability for Enhanced Monitoring of AI Applications

Amazon has introduced CloudWatch generative AI observability (Preview), a specialized monitoring solution for generative AI applications across various environments, including Amazon Bedrock and on-premises infrastructure. This tool offers end-to-end tracing, deep insights into performance and health, and integrates with open-source frameworks using OpenTelemetry. It also leverages existing CloudWatch features for comprehensive monitoring, facilitating the transition from experimentation to production for AI workloads.

Streamlined SQL Server Migration with Azure Arc and Database Migration Service

Azure now offers a streamlined SQL Server migration experience through the Azure Arc migration tool, which integrates with the Azure Database Migration service. This solution provides automated assessments, cost visibility, simplified provisioning, and near real-time database replication, all within a single portal. Additionally, Microsoft Copilot assists with decision-making during the migration process, ensuring minimal downtime and readiness validation.

Azure Application Gateway for Containers Introduces Web Application Firewall Support

Azure's Application Gateway for Containers now supports Web Application Firewall (WAF) in public preview, enhancing protection for Kubernetes workloads against web-based attacks such as SQL injections and cross-site scripting. Users gain access to Azure-managed Default Rulesets for OWASP threats, bot protection through manager rulesets, and DDoS attack mitigation via rate limiting custom rules.

Azure Database for MySQL Introduces Configurable Backup Intervals for Faster Restores

Azure Database for MySQL now allows users to configure the backup interval for automatic backups, enhancing restore speed. This feature introduces more frequent snapshots, which reduces the number of binlogs that need to be replayed during point-in-time restores, thereby minimizing overall restore time.

Google Cloud Enhances Agent2Agent Protocol with New Features and Deployment Options

Google Cloud has launched version 0.3 of the Agent2Agent (A2A) protocol, enhancing developer capabilities with features like gRPC support and security card signing. The update, which is gaining traction among over 150 organizations, includes native A2A support in the Agent Development Kit (ADK) for easier agent creation. Developers can deploy A2A agents using Agent Engine, Cloud Run, or Google Kubernetes Engine (GKE), with $300 in free credits available for new customers to build proof of concepts.

Veo 3 and Veo 3 Fast Now Generally Available on Vertex AI for Enhanced Video Creation

Google Cloud's Vertex AI has made Veo 3, its most advanced video generation model, generally available, following significant adoption with over 70 million videos created. Veo 3 Fast, designed for rapid video creation, is also now available. Upcoming features include image-to-video capabilities, allowing users to create videos from static images and text prompts. Companies like Canva and BarkleyOKRP are leveraging these advancements for enhanced video production and marketing efficiency.

❤️ You might also like

  • Microsoft becomes the second company to reach $4 trillion market cap LINK

  • Google's new AI acts as a virtual satellite (AlphaEarth Foundations) LINK

  • China summoned Nvidia over H20 chip security risks LINK

  • OpenAI is building Stargate Norway while its annual spending soaring to $8 billion LINK

  • OpenAI launches Study Mode for ChatGPT LINK

  • YouTube will use AI to spot teen accounts via age estimation LINK

  • Mark Zuckerberg promises you can trust him with superintelligent AI LINK

  • Microsoft Edge transforms into an AI browser with Copilot Mode LINK

🏁 Enjoy this newsletter?

Forward it to a friend, and let them know they can subscribe here.