Why should you care?

Data lakes hold CSV/Parquet; analysts want T-SQL joins without copying everything into DataVerse first.

See it live — copy this example

Run in SQL Server Management Studio (SSMS) or Azure Data Studio.

-- Conceptual external table pattern
-- CREATE EXTERNAL DATA SOURCE ...
-- CREATE EXTERNAL FILE FORMAT ...
-- CREATE EXTERNAL TABLE dbo.ExtSales (...) WITH (LOCATION='...', DATA_SOURCE=..., FILE_FORMAT=...);
-- SELECT TOP (100) * FROM dbo.ExtSales;
SELECT name, type_desc FROM sys.external_tables;

What happened?

• Listing external tables shows what is already virtualized.
• Creating sources/formats is environment-specific; the SELECT pattern is how you consume them.

Practice next

1. Check if PolyBase/external tables are available in your edition.
2. Start with a small CSV in blob/ADLS.
3. Query TOP (100) before joining huge sets.
4. CTAS-like INSERT…SELECT from external into FactOrderDaily.
5. Compare query times external vs imported.

Remember

Virtualize external data via external tables. Great for lake exploration. Materialize what must be fast/OLTP.

Lake + SQL mashup

DataVerse joins product dims to lake clickstream external table.

Outcome: Analysts prototype without waiting on full ETL.