Microsoft Fabric Blog provides a detailed summary of October 2025 platform updates, focusing on enhanced security, smarter data engineering, new integrations, and workflow improvements for data professionals.

Fabric October 2025 Feature Summary

This October 2025 update for Microsoft Fabric introduces a comprehensive range of improvements across the platform, empowering data engineers, analysts, and IT professionals with new capabilities for security, efficiency, and collaboration.

Events & Announcements

  • Fabric Data Days: Two months of community-driven learning and networking for data and AI professionals, beginning November 4th.
  • FabCon 2026: Annual community event focused on Power BI, Microsoft Fabric SQL, Real-Time Intelligence, and more.

Platform Enhancements

Enhanced Security

  • Outbound Access Protection for Spark (GA): Tightens governance of outbound Spark connections, reducing risk of data exfiltration and expanding granular controls for organizations.
  • Workspace-Level Private Link (GA): Enables secure, private access to Fabric workspaces from internal networks, eliminating public internet exposure.

OneLake Security & Collaboration

  • ReadWrite Access in OneLake: Data owners can assign precise read/write permissions, supporting collaborative workflows without sacrificing governance. Access is managed using Spark notebooks, OneLake File Explorer, or APIs and enforces least privilege.

Data Engineering Innovation

  • Adaptive Target File Size: File size targets in Delta Tables adjust dynamically as table size changes, optimizing for performance, parallelism, and data skipping.
  • Fast Optimize & Auto Compaction: Automated compaction optimizes storage efficiency, reduces write amplification, and minimizes performance dips between maintenance cycles.
  • Spark Connector for SQL Databases (Preview): Native Spark-to-SQL connector with PySpark and Scala support, seamless authentication, and built-in security enforcement.
  • Spark Executor Rolling Logs: Simplifies log management for long or large Spark jobs by segmenting logs hourly for easier troubleshooting.

Improved Developer Experience

  • Keyboard Shortcuts & Focus Mode: Enhanced UI multitasking and distraction-free editing for developers, similar to professional IDEs.
  • Table Deep Links in Lakehouse Explorer: Direct URLs to specific tables facilitate fast preview and collaboration.
  • Fabric Connection with Notebook: Streamlines external data access and credential management in connected notebooks.

Data Agent and Science Updates

  • Data Agent Integration in Lakehouse: Integrate and manage Data Agents directly within Lakehouse, streamlining analytics workflows and enabling AI-driven insights.
  • Markdown Editor for Data Agent Creators: Compose agent instructions in Markdown to improve clarity and LLM understanding.
  • New Authoring Experience: Split Data/Setup tabs, integration with CI/CD, and schema-exploration enhance agent creation and management.

Data Warehouse Advancements

  • Ingesting JSONL Files: OPENROWSET in T-SQL supports reading and transforming JSON Lines data at scale.
  • Data Source Option for OPENROWSET: Enables the use of relative file paths and simplifies access to Lakehouse files.
  • Improved Concurrency – Compaction Preemption: System awareness of active workloads helps avoid write-write conflicts and query failures during compaction.

Real-time Intelligence

  • Native Graph Engine: Horizontally scalable graph data management and analytics inspired by LinkedIn’s system, foundational for agent-driven intelligence.
  • Eventstream: New CDC support for MongoDB, schema validation for EventHub sources, and pause/resume controls for derived streams improve real-time pipeline flexibility and governance.

Data Factory & Power BI Enhancements

  • Copy Job: Now supports ORC, Excel, Avro, and XML formats as well as advanced CSV parsing features for broader interoperability.
  • Variable Libraries in Dataflow Gen2: Gateway support and improved editor integration streamline solution authoring.
  • Export Query Results (Preview): Power BI Desktop users can export data directly to Fabric destinations, supporting seamless cross-product data movement and workflow continuity.

References & Further Reading

This release continues Microsoft Fabric’s vision for a unified, intelligent, and secure analytics platform.

This post appeared first on “Microsoft Fabric Blog”. Read the entire article here