Features

Explore Airport’s key features: SQL operations on remote data, time travel queries, custom functions, row-level security, and Apache Arrow-based high-performance data access.

The Airport extension brings Arrow Flight support to DuckDB, enabling high-performance remote data access and distributed operations. Key features are detailed below.

Key Features

Supported SQL Operations

Databases:

  • ATTACH — Connect remote databases
  • DETACH — Disconnect databases

Tables:

Schemas:

Time Travel

Airport supports time travel queries on tables where the remote system provides this capability. Query historical data snapshots or audit changes over time using standard SQL syntax.

Custom Scalar Functions (UDFs)

Airport supports custom scalar functions that execute on the remote server. These user-defined functions are registered in the remote schema and behave like native DuckDB functions within queries.

Table-Returning Functions

Airport supports table-valued functions (TVFs) that accept arguments—including other tables—and return result sets. TVFs enable parameterized queries, dynamic filtering, and complex data pipelines with remote execution.

Developer Features

Open Source and MIT Licensed

Airport is freely available under the permissive MIT License.

Built on Apache Arrow

Airport leverages Apache Arrow and Arrow Flight for:

  • High-performance inter-process communication (IPC)
  • Efficient streaming across process boundaries
  • Multi-language compatibility (Python, Rust, Go, C++, Java, and more)

Easy to Extend

Airport simplifies the creation of custom Arrow Flight servers with DuckDB-compatible semantics. Connect your data source, expose it via Arrow Flight, and enable SQL queries—no complex integration required.

Row and Column Level Security

Airport supports fine-grained access controls, including:

  • Row-level filtering based on user or session context
  • Column-level redaction or exclusion
  • Integration with secret managers and policy engines