​
Multi-Engine

xorq’s multi-engine system enables seamless data movement between different query engines, allowing you to leverage the strengths of each engine while maintaining a unified workflow.

​
The into_backend Operator

The core of xorq’s multi-engine capability is the into_backend operator, which enables:

  • Transparent data movement between engines
  • Zero-copy data transfer using Apache Arrow
  • Automatic optimization of data placement
import xorq as xo
from xorq.expr.relations import into_backend

# Connect to different engines
pg = xo.postgres.connect_env()
db = xo.duckdb.connect()

# Get tables from different sources
batting = pg.table("batting")

# Load awards_players into DuckDB
awards_players = xo.examples.awards_players.fetch(backend=db)

# Filter data in respective engines
left = batting.filter(batting.yearID == 2015)
right = awards_players.filter(awards_players.lgID == "NL").drop("yearID", "lgID")

# Move right table into postgres for efficient join
expr = left.join(
    into_backend(right, pg),
    ["playerID"],
    how="semi"
)[["yearID", "stint"]]

# Execute the multi-engine query
result = expr.execute()

​
Supported Engines

xorq currently supports:

  1. In-Process Engines
    • DuckDB
    • DataFusion
    • Pandas
  2. Distributed Engines
    • Trino
    • Snowflake
    • BigQuery

​
Engine Selection Guidelines

Choose engines based on their strengths:

  1. DuckDB: Local processing, AsOf joins, efficient file formats
  2. DataFusion: Custom UDFs, streaming processing
  3. Trino: Distributed queries, federation, security
  4. Snowflake/BigQuery: Managed infrastructure, scalability

​
Data Transfer

Data movement between engines is handled through:

  1. Arrow Flight: Zero-copy data transfer protocol
  2. Memory Management: Automatic spilling to disk
  3. Batching: Efficient chunk-based processing