Data Virtualization Demos

A collection of demos showcasing Data Virtualization capabilities in SQL: query external data sources such as Parquet and CSV directly using OPENROWSET, external tables, and related features without moving data.

Demos

Demo	Platform	Description
HammerDB TPROC-H SF10 data virtualization	SQL Server 2025 and Azure SQL Database	Public TPROC-H-generated SF10 backups, BACPACs, and anonymous Parquet data with restore/import instructions, validation, and sample workload results. This is not an official TPC-H data set or benchmark.
Seattle Safety	Azure SQL Database	End-to-end walkthrough using a public 911 dispatch dataset: ad-hoc exploration with OPENROWSET, schema discovery, filepath()/filename(), data ingestion, external tables, hybrid queries, geospatial analysis, and hot/cold data tiering.

Getting Started

1. Choose a demo from the catalog and follow its README.
2. Provision the SQL target described by that demo.
3. Run the scripts in the documented order.
4. All source data is public; storage credentials are not required for reads.

Official Microsoft documentation

• Connect, query, and export data with PolyBase
• PolyBase overview
• CREATE EXTERNAL DATA SOURCE
• CREATE EXTERNAL FILE FORMAT
• CREATE EXTERNAL TABLE

Object naming convention

The Seattle Safety scripts use a consistent suffix convention so each object's tier is obvious:

Suffix	Meaning
_Local	Local table (fully ingested)
_External	External table (remote Parquet, full dataset)
_Hot	Local table (recent data, indexed)
_Cold	External table (historical data, remote)
vw_..._Cold	OPENROWSET view over cold data (supports filepath() partition elimination)
vw_..._All	Unified view combining hot + cold