Author: Twinkle Cyril - Senior Product Manager

___________________________________________________

As your data warehouse evolves with changing business needs, so does your schema. Whether you're onboarding new data sources, updating business logic, or scaling analytics models, schema updates—such as increasing column length or adjusting numeric precision are a normal part of operating a modern analytical warehouse.

Now, even minor schema changes often require rebuilding tables and coordinating downstream deployments. A change as small as expanding a VARCHAR column can turn into a full operational effort impacting ingestion pipelines, CI/CD deployments, and reporting dependencies.

Now, we’re introducing support for ALTER TABLE … ALTER COLUMN in Microsoft Fabric Data Warehouse (Preview), enabling supported schema changes directly on existing warehouse tables using familiar T‑SQL syntax.

Evolve your schema without rewriting data

With ALTER COLUMN support in Fabric Data Warehouse, you can now make supported changes to column definitions without requiring full table rebuilds or rewriting underlying Parquet data files.

Capabilities:

• Expand column sizes as business requirements grow.
• Adjust numeric precision to reflect evolving calculations.
• Modify supported data types in place.
• Update schemas without breaking deployment pipelines.
• Maintain compatibility with downstream queries and reports.

All while continuing to use the same, familiar T‑SQL experience.

Why this matters for analytics teams

Schema evolution is one of the most disruptive operational tasks in analytical environments.

Traditionally in Fabric Data Warehouse, making even minor structural changes to warehouse tables often involves:

• Creating replacement tables.
• Copying existing data using CTAS.
• Reconfiguring ingestion pipelines.
• Updating dependent reports or semantic models.

These workflows introduce deployment delays and increase the risk of inconsistencies across analytics workloads.

ALTER COLUMN support helps reduce this operational overhead by allowing supported schema updates to be applied directly helping teams respond faster to changing business requirements without turning every schema change into a data migration project.

Metadata‑only schema changes for supported scenarios

For supported operations in this preview release, ALTER COLUMN updates the table’s metadata without modifying stored data files.

As a result:

• Existing data ingestion pipelines can continue uninterrupted.
• Stored data remains unchanged.
• Supported schema updates complete significantly faster.
• Operational effort associated with full table rebuilds is reduced.

This enables faster iteration on warehouse schemas while maintaining compatibility with existing workloads.

Supported schema evolution scenarios

In this preview release, ALTER COLUMN supports metadata‑only schema evolution where the updated column definition remains compatible with existing stored data.

Supported scenarios include:

• Increasing integer sizes.
• Increasing floating‑point precision.
• Expanding decimal precision or scale.
• Expanding string or binary column lengths.
• Increasing datetime precision.
• Interchangeable SQL numeric types (such as DECIMAL and NUMERIC).

These changes can be applied directly without requiring underlying data validation or file rewrites.

For a full list of supported conversions, refer to ALTER TABLE (Transact-SQL).

Considerations for cross‑engine compatibility

Some ALTER COLUMN operations may enable type widening at the storage layer.

When this occurs, external engines accessing the same Delta tables must support compatible read‑time type interpretation.

Type widening is described in the delta documentation delta/PROTOCOL.md at master · delta-io/delta, and occurs when an existing data type is compatible with a wider type.

Customers who need to remove type widening from a table schema can create a new table using:

CREATE TABLE AS SELECT (CTAS)

Refer to Delta Lake table format interoperability to learn more about engines that support type widening.

Get started today

You can begin using ALTER COLUMN today in Microsoft Fabric Data Warehouse to apply supported schema updates using familiar T‑SQL syntax—without requiring full table rebuilds.

Support for additional schema evolution scenarios will continue to expand as the feature progresses beyond Preview.

Please refer to the documentation, ALTER TABLE (Transact-SQL),