Period: today · Items: 4 · Source: Azure official updates RSS
Today’s Azure updates show a fairly clear direction. Modernization is becoming smarter, while data is being moved less and read and written more directly. In particular, the combination of Azure Migrate and GitHub Copilot Modernization, along with stronger integration between Azure Databricks and Microsoft OneLake, stands out as news that could prompt practitioners to redraw their architectures.
· Application modernization — When portfolio-level assessment is connected with code-level analysis, migration prioritization becomes far more concrete.
· OneLake — It is worth understanding the concept of unified storage, where multiple engines use the same data layer without copying data.
· Unity Catalog — This is a good opportunity to look at how Databricks integrates data access control and metadata management.
· Delta tables — Where managed Delta tables are stored is directly tied to operational complexity and governance.
· ICMP on NAT Gateway — Even basic diagnostics such as “can it ping?” can vary depending on service capabilities in network operations.
1 item
· This capability combines Azure Migrate’s portfolio-level discovery/assessment with GitHub Copilot Modernization’s context-aware code analysis.
· The key point is that application modernization reviews can now extend beyond a purely infrastructure perspective to include code-level insights.
· When deciding “what should we move first?”, this points to a workflow that considers not just the application inventory, but also code change complexity and modernization potential.
· For organizations with large application portfolios, precision in the initial assessment stage is critical, making this especially relevant in practice.
What it is: This is a preview feature that adds GitHub Copilot Modernization integration to Azure Migrate to support at scale code assessments. What is new is that it connects traditional asset discovery and assessment with context-based code analysis.
Why it matters: Migration and modernization do not end with infrastructure relocation; ultimately, the cost of code changes often determines the outcome. This feature helps close that gap. From an engineering perspective, it provides a more evidence-based way to decide whether an application should be rehosted or refactored.
Try it: First review the Azure Migrate assessment flow, then map out—based on the documentation—how the journey from “portfolio assessment → code assessment” connects in the context of GitHub Copilot Modernization.
Source: https://azure.microsoft.com/updates?id=566145
2 items
· Azure Databricks can now natively read data in Microsoft OneLake through Unity Catalog.
· The key point is that OneLake data can be queried and analyzed without moving or copying the data (query without copy).
· This can reduce duplicate data storage, simplify analytics paths, and improve access speed.
· This is especially noteworthy for organizations using both Databricks and Microsoft Fabric/OneLake.
What it is: Native reading of data stored in OneLake from Azure Databricks using Unity Catalog is now GA. The key point is the ability to use OneLake directly as a data source, rather than assuming separate replication or movement.
Why it matters: One of the most frustrating problems in data platform design is having to copy the same data across multiple storage systems, and this kind of integration makes the architecture much cleaner. Engineers can view Databricks and the Fabric ecosystem not as competing platforms, but as multiple engines on top of a shared storage layer.
Try it: As you read through the documentation on how OneLake and Unity Catalog connect, compare the architectural differences between a “copy-based pipeline” and a “direct-read analytics” model in a diagram.
Source: https://azure.microsoft.com/updates?id=565733
· This time, beyond reading, Azure Databricks can now natively store managed Delta tables in Microsoft OneLake.
· In other words, OneLake can now serve as the storage layer for Databricks workloads, reducing the burden of managing separate storage accounts.
· The vision of OneLake as a unified storage layer has become much clearer.
· With read access reaching GA and write support entering Preview, this suggests Databricks–OneLake integration is expanding quickly.
What it is: This is a preview feature that allows Azure Databricks to write managed Delta tables directly to OneLake. It means there is now an option to make OneLake the core Databricks storage location itself, rather than treating it as a simple integration target.
Why it matters: In data engineering, standardizing storage affects operational complexity, governance, and cost structure. Understanding this feature makes it easier to think in terms of “Databricks as the compute/analytics engine, OneLake as the shared data layer.”
Try it: After organizing the roles of managed Delta tables, Unity Catalog, and OneLake, compare the operational differences between “Databricks-dedicated storage” and “OneLake-integrated storage” in a table.
Source: https://azure.microsoft.com/updates?id=565706
1 item
· Azure Standard V2 NAT Gateway now supports outbound ICMP Echo Request / Echo Reply.
· Workloads behind Standard V2 NAT Gateway can now quickly verify outbound connectivity with tools such as ping.
· From an operations perspective, this effectively adds one of the most basic diagnostic methods for handling network issues.
· The feature may look small, but its practical value in troubleshooting is significant.
What it is: ICMP echo traffic support in Standard V2 NAT Gateway is now generally available. This makes it easier and more intuitive to validate connectivity for workloads behind NAT, where there were previously limitations.
Why it matters: In networking, these “small but practical” improvements often have a more direct impact on operational productivity than major headline features. For those studying Azure networking, this is a reminder to look not only at the role of NAT Gateway, but also at which protocols are actually supported in practice.
Try it: As you review Azure NAT Gateway architecture, create a diagram showing what outbound paths are formed for VMs/workloads behind Standard V2 NAT Gateway.
Source: https://azure.microsoft.com/updates?id=565487
There are no separate retirement items in today’s list.