Write from Dataflow to Apache Iceberg

To write from Dataflow to Apache Iceberg, use the managed I/O connector.

Managed I/O supports the following capabilities for Apache Iceberg:

Catalogs	Hadoop Hive REST-based catalogs BigQuery metastore (requires Apache Beam SDK 2.62.0 or later if not using Runner v2)
Read capabilities	Batch read
Write capabilities	Batch write Streaming write Dynamic destinations Dynamic table creation

For BigQuery tables for Apache Iceberg, use the BigQueryIO connector with BigQuery Storage API. The table must already exist; dynamic table creation is not supported.

Dependencies

Add the following dependencies to your project:

Java

<dependency>
  <groupId>org.apache.beam</groupId>
  <artifactId>beam-sdks-java-managed</artifactId>
  <version>${beam.version}</version>
</dependency>

<dependency>
  <groupId>org.apache.beam</groupId>
  <artifactId>beam-sdks-java-io-iceberg</artifactId>
  <version>${beam.version}</version>
</dependency>

Configuration

Managed I/O uses the following configuration parameters for Apache Iceberg:

Read and write configuration	Data type	Description
table	string	The identifier of the Apache Iceberg table. Example: "db.table1".
catalog_name	string	The name of the catalog. Example: "local".
catalog_properties	map	A map of configuration properties for the Apache Iceberg catalog. The required properties depend on the catalog. For more information, see CatalogUtil in the Apache Iceberg documentation.
config_properties	map	An optional set of Hadoop configuration properties. For more information, see CatalogUtil in the Apache Iceberg documentation.
Write configuration	Data type	Description
triggering_frequency_seconds	integer	For streaming write pipelines, the frequency at which the sink attempts to produce snapshots, in seconds.

Dynamic destinations

Managed I/O for Apache Iceberg supports dynamic destinations. Instead of writing to a single fixed table, the connector can dynamically select a destination table based on field values within the incoming records.

To use dynamic destinations, provide a template for the table configuration parameter. For more information, see Dynamic destinations.

Examples

The following examples show how to use Managed I/O to write to Apache Iceberg.

Write to an Apache Iceberg table

The following example writes in-memory JSON data to an Apache Iceberg table.

Java

To authenticate to Dataflow, set up Application Default Credentials. For more information, see Set up authentication for a local development environment.

import com.google.common.collect.ImmutableMap;
import java.util.Arrays;
import java.util.List;
import java.util.Map;
import org.apache.beam.sdk.Pipeline;
import org.apache.beam.sdk.managed.Managed;
import org.apache.beam.sdk.options.Description;
import org.apache.beam.sdk.options.PipelineOptions;
import org.apache.beam.sdk.options.PipelineOptionsFactory;
import org.apache.beam.sdk.schemas.Schema;
import org.apache.beam.sdk.transforms.Create;
import org.apache.beam.sdk.transforms.JsonToRow;
import org.apache.beam.sdk.values.PCollectionRowTuple;

public class ApacheIcebergWrite {
  static final List<String> TABLE_ROWS = Arrays.asList(
      "{\"id\":0, \"name\":\"Alice\"}",
      "{\"id\":1, \"name\":\"Bob\"}",
      "{\"id\":2, \"name\":\"Charles\"}"
  );

  static final String CATALOG_TYPE = "hadoop";

  // The schema for the table rows.
  public static final Schema SCHEMA = new Schema.Builder()
      .addStringField("name")
      .addInt64Field("id")
      .build();

  public interface Options extends PipelineOptions {
    @Description("The URI of the Apache Iceberg warehouse location")
    String getWarehouseLocation();

    void setWarehouseLocation(String value);

    @Description("The name of the Apache Iceberg catalog")
    String getCatalogName();

    void setCatalogName(String value);

    @Description("The name of the table to write to")
    String getTableName();

    void setTableName(String value);
  }

  public static void main(String[] args) {

    // Parse the pipeline options passed into the application. Example:
    //   --runner=DirectRunner --warehouseLocation=$LOCATION --catalogName=$CATALOG \
    //   --tableName= $TABLE_NAME
    // For more information, see https://beam.apache.org/documentation/programming-guide/#configuring-pipeline-options
    Options options = PipelineOptionsFactory.fromArgs(args).withValidation().as(Options.class);
    Pipeline pipeline = Pipeline.create(options);

    // Configure the Iceberg source I/O
    Map catalogConfig = ImmutableMap.<String, Object>builder()
        .put("warehouse", options.getWarehouseLocation())
        .put("type", CATALOG_TYPE)
        .build();

    ImmutableMap<String, Object> config = ImmutableMap.<String, Object>builder()
        .put("table", options.getTableName())
        .put("catalog_name", options.getCatalogName())
        .put("catalog_properties", catalogConfig)
        .build();

    // Build the pipeline.
    pipeline.apply(Create.of(TABLE_ROWS))
        .apply(JsonToRow.withSchema(SCHEMA))
        .apply(Managed.write(Managed.ICEBERG).withConfig(config));

    pipeline.run().waitUntilFinish();
  }
}

Write with dynamic destinations

The following example writes to different Apache Iceberg tables based on a field in the input data.

Java

To authenticate to Dataflow, set up Application Default Credentials. For more information, see Set up authentication for a local development environment.

import com.google.common.collect.ImmutableMap;
import java.util.Arrays;
import java.util.List;
import java.util.Map;
import org.apache.beam.sdk.Pipeline;
import org.apache.beam.sdk.PipelineResult;
import org.apache.beam.sdk.managed.Managed;
import org.apache.beam.sdk.options.Description;
import org.apache.beam.sdk.options.PipelineOptions;
import org.apache.beam.sdk.options.PipelineOptionsFactory;
import org.apache.beam.sdk.schemas.Schema;
import org.apache.beam.sdk.transforms.Create;
import org.apache.beam.sdk.transforms.JsonToRow;

public class ApacheIcebergDynamicDestinations {

  // The schema for the table rows.
  public static final Schema SCHEMA = new Schema.Builder()
      .addInt64Field("id")
      .addStringField("name")
      .addStringField("airport")
      .build();

  // The data to write to table, formatted as JSON strings.
  static final List<String> TABLE_ROWS = List.of(
      "{\"id\":0, \"name\":\"Alice\", \"airport\": \"ORD\" }",
      "{\"id\":1, \"name\":\"Bob\", \"airport\": \"SYD\" }",
      "{\"id\":2, \"name\":\"Charles\", \"airport\": \"ORD\" }"
  );

  public interface Options extends PipelineOptions {
    @Description("The URI of the Apache Iceberg warehouse location")
    String getWarehouseLocation();

    void setWarehouseLocation(String value);

    @Description("The name of the Apache Iceberg catalog")
    String getCatalogName();

    void setCatalogName(String value);
  }

  // Write JSON data to Apache Iceberg, using dynamic destinations to determine the Iceberg table
  // where Dataflow writes each record. The JSON data contains a field named "airport". The
  // Dataflow pipeline writes to Iceberg tables with the naming pattern "flights-{airport}".
  public static void main(String[] args) {
    // Parse the pipeline options passed into the application. Example:
    //   --runner=DirectRunner --warehouseLocation=$LOCATION --catalogName=$CATALOG \
    // For more information, see https://beam.apache.org/documentation/programming-guide/#configuring-pipeline-options
    Options options = PipelineOptionsFactory.fromArgs(args).withValidation().as(Options.class);
    Pipeline pipeline = Pipeline.create(options);

    // Configure the Iceberg source I/O
    Map catalogConfig = ImmutableMap.<String, Object>builder()
        .put("warehouse", options.getWarehouseLocation())
        .put("type", "hadoop")
        .build();

    ImmutableMap<String, Object> config = ImmutableMap.<String, Object>builder()
        .put("catalog_name", options.getCatalogName())
        .put("catalog_properties", catalogConfig)
        // Route the incoming records based on the value of the "airport" field.
        .put("table", "flights-{airport}")
        // Specify which fields to keep from the input data.
        .put("keep", Arrays.asList("name", "id"))
        .build();

    // Build the pipeline.
    pipeline
        // Read in-memory JSON data.
        .apply(Create.of(TABLE_ROWS))
        // Convert the JSON records to Row objects.
        .apply(JsonToRow.withSchema(SCHEMA))
        // Write each Row to Apache Iceberg.
        .apply(Managed.write(Managed.ICEBERG).withConfig(config));

    // Run the pipeline.
    pipeline.run().waitUntilFinish();
  }
}

What's next

• Read from Apache Iceberg.
• Learn more about Managed I/O.