Chris Romp

We have discovered that an issue can occur with #sqlserver2025 if you have a large number of databases for your instance. We have documented this problem at https://lnkd.in/g-SmMk3r. On this page you will see a workaround to use a startup trace flag to avoid the problem. You are fully supported in production to use this until we produce a fix in a future cumulative update. Feel free to comment if you have questions about this. #sqlserver

318

21 Comments

🚀 We're expanding model choice by integrating Anthropic’s Claude Sonnet 4 and Claude Opus 4.1 into Microsoft 365 Copilot and Microsoft Copilot Studio!. This means more flexibility and power for users—whether you're diving deep with the Researcher agent or building enterprise-grade agents. Now you can mix and match models from OpenAI, Anthropic, and the Azure Model Catalog to tailor your AI experience to your business needs. This is a huge step forward in making Copilot even more adaptable, intelligent, and impactful. 🔍 Learn more about how this unlocks new possibilities for reasoning, automation, and agent orchestration. #Microsoft365 #Copilot #AI #Anthropic #Claude #CopilotStudio #PowerPlatform #FutureOfWork

4

If you haven't checked out DQX for Data Quality, you need to read this. 🎉 DQX v0.9.3 is all about boosting your data quality game with better scalability, flexibility, and usability. Check out the highlights: 🚀 Multi-Table Checks: Run checks across multiple tables in a single go! 🗺️ New Validations: Introducing checks for IPv6 addresses, comprehensive Spatial Data (geometry, OGC validity), and ensuring your DataFrame conforms to the expected Schema! 📐 🧐 Smarter Comparisons: Enhanced dataset comparisons now include absolute and relative tolerance. 🎯 Targeted Profiling: Use SQL filters to profile specific subsets of your data for more relevant insights. 🛠️ Flexible Installs: Install DQX in custom workspaces and use group-assigned clusters. Ready to level up your data quality? Check out DQX v0.9.3! 👇 #DataQuality #DQX #DataEngineering #NewRelease

5

Microsoft has released .NET Aspire 9.5, the latest update to its framework for building distributed cloud-native applications in .NET. Key highlights include: New single-file AppHost option Generative AI Visualizer in the dashboard Expanded AI integrations with GitHub Models, Azure AI Foundry, and OpenAI Lifecycle hooks with startup/readiness/liveness probes Unified multi-resource logging Read the announcement:

1

Power is for Open Source Databases too ! See how MariaDB on IBM Power allows you to resize CPU and memory in real time—without taking your database offline. This video demo shows the power of dynamic scaling in action, giving your workloads the flexibility to grow (or shrink) based on demand—all with zero disruption. https://lnkd.in/g3z5k6Qt

15

🚀 New AI Capabilities in Azure Database for PostgreSQL! If you’re exploring ways to bring Generative AI directly into your database workflows, check out the newly released semantic operators in Azure PostgreSQL Flexible Server. These functions allow you to integrate LLM-powered intelligence directly in your SQL queries—no need for separate pipelines! Here’s what you can now do natively in PostgreSQL with azure_ai: 🔹 azure_ai.generate() – Generate text or structured output using LLMs 🔹 azure_ai.is_true() – Evaluate the likelihood of a statement being true 🔹 azure_ai.extract() – Extract entities or structured data from raw text 🔹 azure_ai.rank() – Rerank documents based on semantic relevance to a query 📖 Learn more: https://lnkd.in/e_3xapFs #PostgreSQL #AzureDatabase #GenerativeAI #LLM #IntelligentApps #SemanticSearch #AzureAI

30

**.NET 10's Memory Revolution: Less Heap, Smarter GC, Faster Cloud Apps** .NET 10 LTS just dropped at .NET Conf and Ignite 2025—perfect for Azure-first workloads. Key wins for cloud pros: JIT escape analysis stacks delegates, lambdas, and small value-type arrays (no escape = no heap tax). GC's DATAS mode (default now) dynamically sizes heaps to your live data, slashing waste in containers and bursty APIs. Benchmarks? 3x faster lambdas, 70% fewer allocations. As an Azure architect who's tuned 100+ prod services, this is huge—say goodbye to GC pauses killing SLAs on App Service/Functions. No code changes needed; just upgrade. Ideal for #CloudNative AI agents without micro-optimizing LINQ/foreach. Your heap won't balloon on bigger SKUs—pay only for what you use. Toggle DATAS if needed for latency beasts. Read the full article here: https://lnkd.in/g4yqA65n

5

This is super cool! Microsoft has been contributing to the PostgreSQL open-source project since 2019, and just for the upcoming version 18, so far they have 456 commits with Microsoft authors (or co-authors), which is about 18.6% of all commits. Some of these commits include stuff like: - Asynchronous I/O subsystem - OAuth infrastructure for authentication - Early groundwork for Multithreading (yes, it's happening) - Amortize aggressive vacuum cost - Fast database cloning - Reduce planning time for partitioned tables with many partitions - Optimize memory management for materialize, window aggregation, & CTE scans - Speed up hash GROUP BY & Hash Join - Remove redundant GROUP BY columns using UNIQUE indexes - Refactor executor to allow inlining of core logic - etc, see the article for the full list. I wish that more tech giants would contribute even more. They are all benefiting from this project. Also, I wish more people would comprehend the fact that PostgreSQL is not (just) a database. It's a full-blown software development platform. https://lnkd.in/dGq6_yi4

23

4 Comments

I updated my blog post detailing Delta table compatibility nuances between Fabric and Databricks to account for Delta 4.0 functionality: https://lnkd.in/guTwtSZH TL/DR: There's a very high level of combability, so before blindly enabling the new Compatibility Mode on tables in Unity Catalog, make sure to use my Python library to check if there are any table features which may be blocking interop. Once Fabric Runtime 2.0 EPP ships globally, all other read blocking features will be supported (Variants, Type Widening). The Python library `onelake-shortcut-tools` will generate a DataFrame in seconds that details the compatibility based on the target Runtime in Fabric: https://lnkd.in/gTPdji69 Cheers! #DeltaLake #Interoperability #Lakehouse #MicrosoftFabric #FabricSpark #ApacheSpark #Databricks

57

5 Comments

Week 2 of Spring into AI challenged us to write an tool that AI can use. I decided to create a SQL Server DB Profiler skill that an agent can invoke. I wanted to move metrics closer, within my IDE. Usage: "DTUs comin in hot! Check last 3 hours. Show top 3 offenders." Output: A synopsis of top N offending queries over last N hours, invocation counts, cpu waits, and tuning recommendations. I chatted with GPT about what kind of metrics are possible to query SQL Server for. Then I asked it to produce a small suite of sql scripts and a powershell script to map input sentiment to input params in the sql scripts, and aggregate the output into a synopsis that relays the real current state of things, so I can quickly take action. Great for when those monitoring alarms go off.. I mean, of course, they never do, but just in case ;) src: https://lnkd.in/gnRDjsNd https://lnkd.in/gFPY2sAn

9

Bohan Zhang's blog post on how OpenAI has scaled ChatGPT on top of Azure PostgreSQL has generated a lot of commentary on how well PostgreSQL can scale (without sharding). Here is our (Azure PostgreSQL teams) perspective on supporting OpenAI and making PostgreSQL run better in the cloud: https://lnkd.in/gyHtGWgs

27

Part 2 is live: C.R.A.F.T. Framework with .NET and Azure I broke down exactly how we're using systematic AI-assisted development on my current restaurant tech platform. Key insights: ✅ Multi-agent orchestration: Claude Code for complex logic, Copilot for patterns, GPT-5 for troubleshooting ✅ Centralized documentation that all AI agents reference ✅ GitHub Copilot's Azure integration providing real-time infrastructure context ✅ Multi-layer security scanning catching issues automatically ✅ What actually breaks (file sprawl, multi-tenant logic gaps, performance assumptions) The reality: Different AI models excel at different tasks. The teams getting 3-4x velocity aren't using one tool—they're orchestrating multiple agents with consistent documentation and validation. This is production implementation, not proof of concept. Multi-repository architecture. Real security requirements. Actual scaling constraints. Article: https://buff.ly/MPRBRvL What's your experience with multi-agent AI development? #SoftwareDevelopment #ArtificialIntelligence #DotNet #Azure #DevSecOps

2