Apache Iceberg

The open table format for analytic datasets.

            {`// Create an Iceberg table
                        Table table = catalog.createTable(
                        TableIdentifier.of("db", "events"),
                        schema,
                        PartitionSpec.builderFor(schema)
                          .day("timestamp")
                          .build()
                        );

                        // Write data with ACID guarantees
                        table.newAppend()
                        .appendFile(dataFile)
                        .commit();`}

Quick Start

Get up and running with Apache Iceberg in minutes

Add the Iceberg dependency to your build configuration: ```xml Maven theme={null} org.apache.iceberg iceberg-core 1.7.1 ``` ```groovy Gradle theme={null} dependencies { implementation 'org.apache.iceberg:iceberg-core:1.7.1' } ``` Initialize a catalog to manage your tables: ```java theme={null} import org.apache.iceberg.catalog.Catalog; import org.apache.hadoop.conf.Configuration; Configuration conf = new Configuration(); Catalog catalog = new HadoopCatalog(conf, "warehouse/path"); ``` Define a schema and create an Iceberg table: ```java theme={null} import org.apache.iceberg.*; import org.apache.iceberg.types.Types; Schema schema = new Schema( Types.NestedField.required(1, "id", Types.LongType.get()), Types.NestedField.required(2, "data", Types.StringType.get()), Types.NestedField.required(3, "timestamp", Types.TimestampType.withZone()) ); PartitionSpec spec = PartitionSpec.builderFor(schema) .day("timestamp") .build(); Table table = catalog.createTable( TableIdentifier.of("my_db", "my_table"), schema, spec ); ``` Iceberg tables support schema evolution, partitioning evolution, and hidden partitioning to prevent user errors. Use Iceberg tables with Spark, Flink, Trino, or other query engines: ```sql Spark SQL theme={null} -- Read from Iceberg table SELECT * FROM my_db.my_table WHERE timestamp >= current_date(); -- Time travel to previous snapshot SELECT * FROM my_db.my_table VERSION AS OF 12345678901234567890; ``` See the [query engines guide](/engines/spark/overview) for detailed integration instructions.

Why Iceberg?

Built for reliability, performance, and correctness at massive scale

Serializable isolation ensures readers never see partial or uncommitted changes. Multiple concurrent writers use optimistic concurrency. Query historical table snapshots for reproducible analysis and easy version rollback when issues occur. Add, drop, rename, or update columns without side effects. No accidental data deletion or corruption. Users don't need to know about partitioning to get fast queries. Partition layout can evolve as data patterns change. Data files are pruned using partition and column-level statistics from table metadata. Scan planning is fast. Works seamlessly with Spark, Flink, Trino, Presto, Hive, and Impala. All engines can safely work with the same tables.

Explore by Topic

Dive deep into Iceberg's capabilities

Understand the Iceberg table format, schemas, partitioning, and reliability guarantees. Learn how to read, write, and maintain Iceberg tables efficiently. Integrate Iceberg with Spark, Flink, Hive, and other processing engines. Configure and customize catalogs for table metadata management. Use the REST Catalog API for programmatic table management. Migrate existing tables from Hive or Delta Lake to Iceberg.

Ready to Get Started?

Follow our quickstart guide to create your first Iceberg table, or explore the API reference to learn about all available features.

View Quickstart Read Introduction