Query pushdown
Query pushdown is an optimization that evaluates eligible filtering, projection, or aggregation operations as close as possible to the source data.
Architecting an Apache Iceberg Lakehouse
Filter pushdown is a query optimization that evaluates filtering conditions as close as possible to the source data.
- How it is used
- Improves lakehouse, federated, dashboard, and notebook queries by minimizing unnecessary data access.
- How deep it goes
- Explains filter, projection, and aggregation placement, then connects execution to file statistics, row groups, planning, and caching.
- Examples
- Uses WHERE id = 2, Parquet row-group statistics, Iceberg snapshots, federated sources, dashboards, and notebooks as concrete contexts.
- What you can do after
- Apply it via query engines to reduce scanned data and improve latency across Iceberg tables, federated sources, dashboards, and notebooks.
Read how Architecting an Apache Iceberg Lakehouse explains query pushdown →
DuckDB in Action
A pushdown predicate is a query condition moved into an underlying data operation so filtering occurs before later result processing.
- How it is used
- Optimizes analytical queries by filtering Parquet data at the storage layer or beneath a DuckDB view before later processing.
- How deep it goes
- Explains early filtering, Parquet metadata, and view predicate propagation, but stops short of optimizer execution-plan internals.
- Examples
- Demonstrates Parquet conversion, ATP rankings, power-production views, and New York taxi analysis.
- What you can do after
- Apply it via DuckDB to reduce loaded data and network transfer when querying Parquet files or views.
Related concepts
- Parquetused by · Apache Parquet is an open, column-oriented binary file format for storing and processing structured data with schema and metadata support.
- DuckDBused by · DuckDB is an open-source, in-process analytical database management system with a columnar-vectorized execution engine for fast local querying.
- Apache Icebergused by · Apache Iceberg is an open table format for managing large analytical datasets with metadata, schema evolution, snapshots, and transactional operations across distributed or object storage.
- row groupuses · A row group is a storage and processing unit containing a subset of rows in a columnar data file.
- query engineused by · A query engine is a software component that processes queries against data or knowledge sources and produces results.
- consumption layerused by · A lakehouse architectural layer that exposes data to business-intelligence, artificial-intelligence, and real-time applications.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.