Connect to DuckDB (Parquet) through standard ODBC by configuring the ZappySys JDBC Bridge Driver without Java coding.

Download the duckdb_jdbc JAR, create a DSN with a jdbc:duckdb URL (the guide starts with in-memory jdbc:duckdb:memory:), validate with Test Connection, and preview SQL — including read_parquet — before wiring Excel, Power BI, or ETL jobs.
Part of ZappySys ODBC PowerPack (30+ drivers).

Recommended starting point

View Setup Guide & Get Started →

See features, UI & examples →

Pro Tip: Install Java 8+ (Java 21 recommended), download duckdb_jdbc from MVN Repository or the DuckDB Java install page, and point the DSN at a local JAR (for example D:\Drivers\JDBC\duckdb_jdbc.jar). Use a User DSN for interactive design and a System DSN for Windows services; match 32-bit vs 64-bit ODBC Administrator to the host process. For errors after Test Connection, browse the ZappySys Community Parquet-related threads.

Compatibility: Windows ODBC (32/64-bit) | Works with Visual Studio, .NET, BI tools | Data Gateway for SQL Server, JDBC, and non-Windows access

⚡ Key Benefits

Parquet over ODBC — Use DuckDB’s SQL engine (including read_parquet) through a JDBC Bridge DSN your ODBC tools already understand
Simple JDBC footprint — Configure one duckdb_jdbc JAR path in the DSN wizard per the setup guide
Validated connectivity — Use Test Connection plus in-DSN SQL preview before production queries
Operational DSN patterns — Choose User DSN for desktop sessions and System DSN for service accounts (match ODBC bitness)
Optional Data Gateway — Centralize the DuckDB (Parquet) data source for shared ODBC access and simpler control

Learn more about Data Gateway

📸 Visual Overview & Setup

1. Choose JDBC Bridge Driver Type 🔍 View Full Size	2. Enter Full Path for *.jar file (i.e. JDBC driver file) 🔍 View Full Size	3. Preview SQL Query 🔍 View Full Size
4. Load DuckDB (Parquet) Rows into Excel 🔍 View Full Size	5. Connect from Power BI 🔍 View Full Size	6. Share via Data Gateway Data Gateway overview

🔗 Seamless SQL Query Integration

Query Parquet and relational sources with DuckDB SQL

-- Read remote Parquet (example from the setup guide)
SELECT *
FROM read_parquet('https://hrbrmstr.github.io/large-parquet-test/sample.parquet');

✔ Use read_parquet against URLs or local paths
✔ No ETL or custom scripts required for ad hoc reads
✔ Works inside SQL Server, Power BI, Excel

🔧 Key Features

Read & Query

Run DuckDB SQL through ODBC after the DSN is configured, including read_parquet as shown in the guide
Use the preview grid or Select Table metadata picker when the JDBC driver exposes catalogs
See additional patterns in JDBC Bridge SQL query examples linked from the setup guide

JDBC Bridge Configuration

Example JDBC URL from the guide: jdbc:duckdb:memory:
Example DSN name from the guide: DuckdbParquetDSN
Point JDBC driver file(s) at your local JAR (example path: D:\Drivers\JDBC\duckdb_jdbc.jar)
Set Connection parameters to [] when no extra properties are required, or extend per your deployment
Run Test Connection to confirm Java, JAR path, and network reachability (for remote Parquet) before connecting client tools

Writes & Driver Behavior

The bridge forwards SQL unchanged; INSERT/UPDATE/DELETE support depends on your DuckDB JDBC build and database file mode
If you see action is not supported, that reflects the JDBC driver rather than the ZappySys bridge

Reliability & Platform Fit

Standard ODBC 3.x for Power BI, Excel, SQL Server linked servers, SSIS ODBC Source, and custom apps
Java 8 minimum with Java 21 recommended when your JDBC distribution allows it
For Data Gateway deployments, copy DSN settings to the gateway host, adjust firewall rules, and restart the gateway service after edits
Use the guide’s Troubleshooters & resources (JDBC Bridge Driver) list for Java discovery, SSL, multi-JAR merges, logging, and disconnect errors
Learn more about the bridge in ZappySys JDBC Bridge Driver documentation

💡 Common Use Cases

Typical scenarios for the DuckDB (Parquet) ODBC Driver:

Parquet analytics: Let analysts pull remote or on-disk Parquet into Excel or Power BI via one DSN.
Example: prototype with read_parquet against HTTPS-hosted samples.
Staging and integration: Land DuckDB query results into SQL Server for downstream SSIS or reporting.
Example: scheduled extracts using a System DSN under a service account.
Gateway sharing: Publish DuckdbParquetDSN on Data Gateway so ODBC Driver for SQL Server clients connect centrally.
Example: remote teams reuse ZappySysGatewayDSN with default database set to the gateway data source name.
Service automation: Schedule jobs under a System DSN so Windows services and SQL Agent see the same credentials.
Example: align 32-bit services with the 32-bit ODBC administrator.

🎯 Summary

Bring DuckDB (Parquet) into ODBC-native tools without maintaining bespoke JDBC plumbing.
Install Java, add the duckdb_jdbc JAR, build a JDBC Bridge DSN, validate connectivity, and reuse the same DSN across Excel, Power BI, SQL Server, and optional Data Gateway deployments.

Trusted by Developers & IT Teams Worldwide

Built for ODBC Workloads — One driver model across BI, ETL, databases, and custom code.
Expert Technical Support — Engineers help with JDBC URLs, Java installs, and gateway networking.
Proven Enterprise Scale — Thousands of teams worldwide rely on ZappySys for API and cloud connectivity.

ODBC PowerPack

Includes DuckDB (Parquet) ODBC Driver