Apache Iceberg is a new table format for storing large, slow-moving tabular data. It is designed to improve on the de-facto standard table layout built into Hive, Trino, and Spark.
Background and documentation is available at https://iceberg.apache.org
Iceberg is under active development at the Apache Software Foundation.
The core Java library that tracks table snapshots and metadata is complete, but still evolving. Current work is focused on adding row-level deletes and upserts, and integration work with new engines like Flink and Hive.
The Iceberg format specification is being actively updated and is open for comment. Until the specification is complete and released, it carries no compatibility guarantees. The spec is currently evolving as the Java reference implementation changes.
Java API javadocs are available for the master.
Iceberg tracks issues in GitHub and prefers to receive contributions as pull requests.
Community discussions happen primarily on the dev mailing list or on specific issues.
Iceberg is built using Gradle 5.4.1 with Java 1.8 or Java 11.
./gradlew build -x test
Iceberg table support is organized in library modules:
iceberg-commoncontains utility classes used in other modules
iceberg-apicontains the public Iceberg API
iceberg-corecontains implementations of the Iceberg API and support for Avro data files, this is what processing engines should depend on
iceberg-parquetis an optional module for working with tables backed by Parquet files
iceberg-arrowis an optional module for reading Parquet into Arrow memory
iceberg-orcis an optional module for working with tables backed by ORC files
iceberg-hive-metastoreis an implementation of Iceberg tables backed by the Hive metastore Thrift client
iceberg-datais an optional module for working with tables directly from JVM applications
This project Iceberg also has modules for adding Iceberg support to processing engines:
iceberg-spark2is an implementation of Spark's Datasource V2 API in 2.4 for Iceberg (use iceberg-spark-runtime for a shaded version)
iceberg-spark3is an implementation of Spark's Datasource V2 API in 3.0 for Iceberg (use iceberg-spark3-runtime for a shaded version)
iceberg-flinkcontains classes for integrating with Apache Flink (use iceberg-flink-runtime for a shaded version)
iceberg-mrcontains an InputFormat and other classes for integrating with Apache Hive
iceberg-pigis an implementation of Pig's LoadFunc API for Iceberg
Iceberg's Spark integration is compatible with Spark 2.4 and Spark 3.0 using the modules in the following table:
|Iceberg version||Spark 2.4.x||Spark 3.0.x|