All Downloads are FREE. Search and download functionalities are using the official Maven repository.

org.apache.iceberg.flink.sink.FlinkSink Maven / Gradle / Ivy

There is a newer version: 1.7.0
Show newest version
/*
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing,
 * software distributed under the License is distributed on an
 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
 * KIND, either express or implied.  See the License for the
 * specific language governing permissions and limitations
 * under the License.
 */
package org.apache.iceberg.flink.sink;

import static org.apache.iceberg.TableProperties.AVRO_COMPRESSION;
import static org.apache.iceberg.TableProperties.AVRO_COMPRESSION_LEVEL;
import static org.apache.iceberg.TableProperties.ORC_COMPRESSION;
import static org.apache.iceberg.TableProperties.ORC_COMPRESSION_STRATEGY;
import static org.apache.iceberg.TableProperties.PARQUET_COMPRESSION;
import static org.apache.iceberg.TableProperties.PARQUET_COMPRESSION_LEVEL;
import static org.apache.iceberg.TableProperties.WRITE_DISTRIBUTION_MODE;

import java.io.IOException;
import java.io.UncheckedIOException;
import java.time.Duration;
import java.util.List;
import java.util.Map;
import java.util.Set;
import java.util.function.Function;
import org.apache.flink.api.common.functions.MapFunction;
import org.apache.flink.api.common.typeinfo.TypeInformation;
import org.apache.flink.api.common.typeinfo.Types;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.configuration.ReadableConfig;
import org.apache.flink.streaming.api.datastream.DataStream;
import org.apache.flink.streaming.api.datastream.DataStreamSink;
import org.apache.flink.streaming.api.datastream.SingleOutputStreamOperator;
import org.apache.flink.streaming.api.functions.sink.DiscardingSink;
import org.apache.flink.table.api.TableSchema;
import org.apache.flink.table.data.RowData;
import org.apache.flink.table.data.util.DataFormatConverters;
import org.apache.flink.table.types.DataType;
import org.apache.flink.table.types.logical.RowType;
import org.apache.flink.types.Row;
import org.apache.iceberg.DataFile;
import org.apache.iceberg.DistributionMode;
import org.apache.iceberg.FileFormat;
import org.apache.iceberg.PartitionField;
import org.apache.iceberg.PartitionSpec;
import org.apache.iceberg.Schema;
import org.apache.iceberg.SerializableTable;
import org.apache.iceberg.Table;
import org.apache.iceberg.flink.FlinkSchemaUtil;
import org.apache.iceberg.flink.FlinkWriteConf;
import org.apache.iceberg.flink.FlinkWriteOptions;
import org.apache.iceberg.flink.TableLoader;
import org.apache.iceberg.flink.util.FlinkCompatibilityUtil;
import org.apache.iceberg.io.WriteResult;
import org.apache.iceberg.relocated.com.google.common.annotations.VisibleForTesting;
import org.apache.iceberg.relocated.com.google.common.base.Preconditions;
import org.apache.iceberg.relocated.com.google.common.collect.Lists;
import org.apache.iceberg.relocated.com.google.common.collect.Maps;
import org.apache.iceberg.relocated.com.google.common.collect.Sets;
import org.apache.iceberg.types.TypeUtil;
import org.apache.iceberg.util.SerializableSupplier;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

public class FlinkSink {
  private static final Logger LOG = LoggerFactory.getLogger(FlinkSink.class);

  private static final String ICEBERG_STREAM_WRITER_NAME =
      IcebergStreamWriter.class.getSimpleName();
  private static final String ICEBERG_FILES_COMMITTER_NAME =
      IcebergFilesCommitter.class.getSimpleName();

  private FlinkSink() {}

  /**
   * Initialize a {@link Builder} to export the data from generic input data stream into iceberg
   * table. We use {@link RowData} inside the sink connector, so users need to provide a mapper
   * function and a {@link TypeInformation} to convert those generic records to a RowData
   * DataStream.
   *
   * @param input the generic source input data stream.
   * @param mapper function to convert the generic data to {@link RowData}
   * @param outputType to define the {@link TypeInformation} for the input data.
   * @param  the data type of records.
   * @return {@link Builder} to connect the iceberg table.
   */
  public static  Builder builderFor(
      DataStream input, MapFunction mapper, TypeInformation outputType) {
    return new Builder().forMapperOutputType(input, mapper, outputType);
  }

  /**
   * Initialize a {@link Builder} to export the data from input data stream with {@link Row}s into
   * iceberg table. We use {@link RowData} inside the sink connector, so users need to provide a
   * {@link TableSchema} for builder to convert those {@link Row}s to a {@link RowData} DataStream.
   *
   * @param input the source input data stream with {@link Row}s.
   * @param tableSchema defines the {@link TypeInformation} for input data.
   * @return {@link Builder} to connect the iceberg table.
   */
  public static Builder forRow(DataStream input, TableSchema tableSchema) {
    RowType rowType = (RowType) tableSchema.toRowDataType().getLogicalType();
    DataType[] fieldDataTypes = tableSchema.getFieldDataTypes();

    DataFormatConverters.RowConverter rowConverter =
        new DataFormatConverters.RowConverter(fieldDataTypes);
    return builderFor(input, rowConverter::toInternal, FlinkCompatibilityUtil.toTypeInfo(rowType))
        .tableSchema(tableSchema);
  }

  /**
   * Initialize a {@link Builder} to export the data from input data stream with {@link RowData}s
   * into iceberg table.
   *
   * @param input the source input data stream with {@link RowData}s.
   * @return {@link Builder} to connect the iceberg table.
   */
  public static Builder forRowData(DataStream input) {
    return new Builder().forRowData(input);
  }

  public static class Builder {
    private Function> inputCreator = null;
    private TableLoader tableLoader;
    private Table table;
    private TableSchema tableSchema;
    private List equalityFieldColumns = null;
    private String uidPrefix = null;
    private final Map snapshotProperties = Maps.newHashMap();
    private ReadableConfig readableConfig = new Configuration();
    private final Map writeOptions = Maps.newHashMap();
    private FlinkWriteConf flinkWriteConf = null;

    private Builder() {}

    private Builder forRowData(DataStream newRowDataInput) {
      this.inputCreator = ignored -> newRowDataInput;
      return this;
    }

    private  Builder forMapperOutputType(
        DataStream input, MapFunction mapper, TypeInformation outputType) {
      this.inputCreator =
          newUidPrefix -> {
            // Input stream order is crucial for some situation(e.g. in cdc case). Therefore, we
            // need to set the parallelism
            // of map operator same as its input to keep map operator chaining its input, and avoid
            // rebalanced by default.
            SingleOutputStreamOperator inputStream =
                input.map(mapper, outputType).setParallelism(input.getParallelism());
            if (newUidPrefix != null) {
              inputStream.name(operatorName(newUidPrefix)).uid(newUidPrefix + "-mapper");
            }
            return inputStream;
          };
      return this;
    }

    /**
     * This iceberg {@link Table} instance is used for initializing {@link IcebergStreamWriter}
     * which will write all the records into {@link DataFile}s and emit them to downstream operator.
     * Providing a table would avoid so many table loading from each separate task.
     *
     * @param newTable the loaded iceberg table instance.
     * @return {@link Builder} to connect the iceberg table.
     */
    public Builder table(Table newTable) {
      this.table = newTable;
      return this;
    }

    /**
     * The table loader is used for loading tables in {@link IcebergFilesCommitter} lazily, we need
     * this loader because {@link Table} is not serializable and could not just use the loaded table
     * from Builder#table in the remote task manager.
     *
     * @param newTableLoader to load iceberg table inside tasks.
     * @return {@link Builder} to connect the iceberg table.
     */
    public Builder tableLoader(TableLoader newTableLoader) {
      this.tableLoader = newTableLoader;
      return this;
    }

    /**
     * Set the write properties for Flink sink. View the supported properties in {@link
     * FlinkWriteOptions}
     */
    public Builder set(String property, String value) {
      writeOptions.put(property, value);
      return this;
    }

    /**
     * Set the write properties for Flink sink. View the supported properties in {@link
     * FlinkWriteOptions}
     */
    public Builder setAll(Map properties) {
      writeOptions.putAll(properties);
      return this;
    }

    public Builder tableSchema(TableSchema newTableSchema) {
      this.tableSchema = newTableSchema;
      return this;
    }

    public Builder overwrite(boolean newOverwrite) {
      writeOptions.put(FlinkWriteOptions.OVERWRITE_MODE.key(), Boolean.toString(newOverwrite));
      return this;
    }

    public Builder flinkConf(ReadableConfig config) {
      this.readableConfig = config;
      return this;
    }

    /**
     * Configure the write {@link DistributionMode} that the flink sink will use. Currently, flink
     * support {@link DistributionMode#NONE} and {@link DistributionMode#HASH}.
     *
     * @param mode to specify the write distribution mode.
     * @return {@link Builder} to connect the iceberg table.
     */
    public Builder distributionMode(DistributionMode mode) {
      Preconditions.checkArgument(
          !DistributionMode.RANGE.equals(mode),
          "Flink does not support 'range' write distribution mode now.");
      if (mode != null) {
        writeOptions.put(FlinkWriteOptions.DISTRIBUTION_MODE.key(), mode.modeName());
      }
      return this;
    }

    /**
     * Configuring the write parallel number for iceberg stream writer.
     *
     * @param newWriteParallelism the number of parallel iceberg stream writer.
     * @return {@link Builder} to connect the iceberg table.
     */
    public Builder writeParallelism(int newWriteParallelism) {
      writeOptions.put(
          FlinkWriteOptions.WRITE_PARALLELISM.key(), Integer.toString(newWriteParallelism));
      return this;
    }

    /**
     * All INSERT/UPDATE_AFTER events from input stream will be transformed to UPSERT events, which
     * means it will DELETE the old records and then INSERT the new records. In partitioned table,
     * the partition fields should be a subset of equality fields, otherwise the old row that
     * located in partition-A could not be deleted by the new row that located in partition-B.
     *
     * @param enabled indicate whether it should transform all INSERT/UPDATE_AFTER events to UPSERT.
     * @return {@link Builder} to connect the iceberg table.
     */
    public Builder upsert(boolean enabled) {
      writeOptions.put(FlinkWriteOptions.WRITE_UPSERT_ENABLED.key(), Boolean.toString(enabled));
      return this;
    }

    /**
     * Configuring the equality field columns for iceberg table that accept CDC or UPSERT events.
     *
     * @param columns defines the iceberg table's key.
     * @return {@link Builder} to connect the iceberg table.
     */
    public Builder equalityFieldColumns(List columns) {
      this.equalityFieldColumns = columns;
      return this;
    }

    /**
     * Set the uid prefix for FlinkSink operators. Note that FlinkSink internally consists of
     * multiple operators (like writer, committer, dummy sink etc.) Actually operator uid will be
     * appended with a suffix like "uidPrefix-writer". 
*
* If provided, this prefix is also applied to operator names.
*
* Flink auto generates operator uid if not set explicitly. It is a recommended * best-practice to set uid for all operators before deploying to production. Flink has an * option to {@code pipeline.auto-generate-uid=false} to disable auto-generation and force * explicit setting of all operator uid.
*
* Be careful with setting this for an existing job, because now we are changing the operator * uid from an auto-generated one to this new value. When deploying the change with a * checkpoint, Flink won't be able to restore the previous Flink sink operator state (more * specifically the committer operator state). You need to use {@code --allowNonRestoredState} * to ignore the previous sink state. During restore Flink sink state is used to check if last * commit was actually successful or not. {@code --allowNonRestoredState} can lead to data loss * if the Iceberg commit failed in the last completed checkpoint. * * @param newPrefix prefix for Flink sink operator uid and name * @return {@link Builder} to connect the iceberg table. */ public Builder uidPrefix(String newPrefix) { this.uidPrefix = newPrefix; return this; } public Builder setSnapshotProperties(Map properties) { snapshotProperties.putAll(properties); return this; } public Builder setSnapshotProperty(String property, String value) { snapshotProperties.put(property, value); return this; } public Builder toBranch(String branch) { writeOptions.put(FlinkWriteOptions.BRANCH.key(), branch); return this; } private DataStreamSink chainIcebergOperators() { Preconditions.checkArgument( inputCreator != null, "Please use forRowData() or forMapperOutputType() to initialize the input DataStream."); Preconditions.checkNotNull(tableLoader, "Table loader shouldn't be null"); DataStream rowDataInput = inputCreator.apply(uidPrefix); if (table == null) { if (!tableLoader.isOpen()) { tableLoader.open(); } try (TableLoader loader = tableLoader) { this.table = loader.loadTable(); } catch (IOException e) { throw new UncheckedIOException( "Failed to load iceberg table from table loader: " + tableLoader, e); } } flinkWriteConf = new FlinkWriteConf(table, writeOptions, readableConfig); // Find out the equality field id list based on the user-provided equality field column names. List equalityFieldIds = checkAndGetEqualityFieldIds(); // Convert the requested flink table schema to flink row type. RowType flinkRowType = toFlinkRowType(table.schema(), tableSchema); // Distribute the records from input data stream based on the write.distribution-mode and // equality fields. DataStream distributeStream = distributeDataStream( rowDataInput, equalityFieldIds, table.spec(), table.schema(), flinkRowType); // Add parallel writers that append rows to files SingleOutputStreamOperator writerStream = appendWriter(distributeStream, flinkRowType, equalityFieldIds); // Add single-parallelism committer that commits files // after successful checkpoint or end of input SingleOutputStreamOperator committerStream = appendCommitter(writerStream); // Add dummy discard sink return appendDummySink(committerStream); } /** * Append the iceberg sink operators to write records to iceberg table. * * @return {@link DataStreamSink} for sink. */ public DataStreamSink append() { return chainIcebergOperators(); } private String operatorName(String suffix) { return uidPrefix != null ? uidPrefix + "-" + suffix : suffix; } @VisibleForTesting List checkAndGetEqualityFieldIds() { List equalityFieldIds = Lists.newArrayList(table.schema().identifierFieldIds()); if (equalityFieldColumns != null && !equalityFieldColumns.isEmpty()) { Set equalityFieldSet = Sets.newHashSetWithExpectedSize(equalityFieldColumns.size()); for (String column : equalityFieldColumns) { org.apache.iceberg.types.Types.NestedField field = table.schema().findField(column); Preconditions.checkNotNull( field, "Missing required equality field column '%s' in table schema %s", column, table.schema()); equalityFieldSet.add(field.fieldId()); } if (!equalityFieldSet.equals(table.schema().identifierFieldIds())) { LOG.warn( "The configured equality field column IDs {} are not matched with the schema identifier field IDs" + " {}, use job specified equality field columns as the equality fields by default.", equalityFieldSet, table.schema().identifierFieldIds()); } equalityFieldIds = Lists.newArrayList(equalityFieldSet); } return equalityFieldIds; } @SuppressWarnings("unchecked") private DataStreamSink appendDummySink( SingleOutputStreamOperator committerStream) { DataStreamSink resultStream = committerStream .addSink(new DiscardingSink()) .name(operatorName(String.format("IcebergSink %s", this.table.name()))) .setParallelism(1); if (uidPrefix != null) { resultStream = resultStream.uid(uidPrefix + "-dummysink"); } return resultStream; } private SingleOutputStreamOperator appendCommitter( SingleOutputStreamOperator writerStream) { IcebergFilesCommitter filesCommitter = new IcebergFilesCommitter( tableLoader, flinkWriteConf.overwriteMode(), snapshotProperties, flinkWriteConf.workerPoolSize(), flinkWriteConf.branch(), table.spec()); SingleOutputStreamOperator committerStream = writerStream .transform(operatorName(ICEBERG_FILES_COMMITTER_NAME), Types.VOID, filesCommitter) .setParallelism(1) .setMaxParallelism(1); if (uidPrefix != null) { committerStream = committerStream.uid(uidPrefix + "-committer"); } return committerStream; } private SingleOutputStreamOperator appendWriter( DataStream input, RowType flinkRowType, List equalityFieldIds) { // Validate the equality fields and partition fields if we enable the upsert mode. if (flinkWriteConf.upsertMode()) { Preconditions.checkState( !flinkWriteConf.overwriteMode(), "OVERWRITE mode shouldn't be enable when configuring to use UPSERT data stream."); Preconditions.checkState( !equalityFieldIds.isEmpty(), "Equality field columns shouldn't be empty when configuring to use UPSERT data stream."); if (!table.spec().isUnpartitioned()) { for (PartitionField partitionField : table.spec().fields()) { Preconditions.checkState( equalityFieldIds.contains(partitionField.sourceId()), "In UPSERT mode, partition field '%s' should be included in equality fields: '%s'", partitionField, equalityFieldColumns); } } } SerializableTable serializableTable = (SerializableTable) SerializableTable.copyOf(table); Duration tableRefreshInterval = flinkWriteConf.tableRefreshInterval(); SerializableSupplier tableSupplier; if (tableRefreshInterval != null) { tableSupplier = new CachingTableSupplier(serializableTable, tableLoader, tableRefreshInterval); } else { tableSupplier = () -> serializableTable; } IcebergStreamWriter streamWriter = createStreamWriter(tableSupplier, flinkWriteConf, flinkRowType, equalityFieldIds); int parallelism = flinkWriteConf.writeParallelism() == null ? input.getParallelism() : flinkWriteConf.writeParallelism(); SingleOutputStreamOperator writerStream = input .transform( operatorName(ICEBERG_STREAM_WRITER_NAME), TypeInformation.of(WriteResult.class), streamWriter) .setParallelism(parallelism); if (uidPrefix != null) { writerStream = writerStream.uid(uidPrefix + "-writer"); } return writerStream; } private DataStream distributeDataStream( DataStream input, List equalityFieldIds, PartitionSpec partitionSpec, Schema iSchema, RowType flinkRowType) { DistributionMode writeMode = flinkWriteConf.distributionMode(); LOG.info("Write distribution mode is '{}'", writeMode.modeName()); switch (writeMode) { case NONE: if (equalityFieldIds.isEmpty()) { return input; } else { LOG.info("Distribute rows by equality fields, because there are equality fields set"); return input.keyBy( new EqualityFieldKeySelector(iSchema, flinkRowType, equalityFieldIds)); } case HASH: if (equalityFieldIds.isEmpty()) { if (partitionSpec.isUnpartitioned()) { LOG.warn( "Fallback to use 'none' distribution mode, because there are no equality fields set " + "and table is unpartitioned"); return input; } else { return input.keyBy(new PartitionKeySelector(partitionSpec, iSchema, flinkRowType)); } } else { if (partitionSpec.isUnpartitioned()) { LOG.info( "Distribute rows by equality fields, because there are equality fields set " + "and table is unpartitioned"); return input.keyBy( new EqualityFieldKeySelector(iSchema, flinkRowType, equalityFieldIds)); } else { for (PartitionField partitionField : partitionSpec.fields()) { Preconditions.checkState( equalityFieldIds.contains(partitionField.sourceId()), "In 'hash' distribution mode with equality fields set, partition field '%s' " + "should be included in equality fields: '%s'", partitionField, equalityFieldColumns); } return input.keyBy(new PartitionKeySelector(partitionSpec, iSchema, flinkRowType)); } } case RANGE: if (equalityFieldIds.isEmpty()) { LOG.warn( "Fallback to use 'none' distribution mode, because there are no equality fields set " + "and {}=range is not supported yet in flink", WRITE_DISTRIBUTION_MODE); return input; } else { LOG.info( "Distribute rows by equality fields, because there are equality fields set " + "and{}=range is not supported yet in flink", WRITE_DISTRIBUTION_MODE); return input.keyBy( new EqualityFieldKeySelector(iSchema, flinkRowType, equalityFieldIds)); } default: throw new RuntimeException("Unrecognized " + WRITE_DISTRIBUTION_MODE + ": " + writeMode); } } } static RowType toFlinkRowType(Schema schema, TableSchema requestedSchema) { if (requestedSchema != null) { // Convert the flink schema to iceberg schema firstly, then reassign ids to match the existing // iceberg schema. Schema writeSchema = TypeUtil.reassignIds(FlinkSchemaUtil.convert(requestedSchema), schema); TypeUtil.validateWriteSchema(schema, writeSchema, true, true); // We use this flink schema to read values from RowData. The flink's TINYINT and SMALLINT will // be promoted to // iceberg INTEGER, that means if we use iceberg's table schema to read TINYINT (backend by 1 // 'byte'), we will // read 4 bytes rather than 1 byte, it will mess up the byte array in BinaryRowData. So here // we must use flink // schema. return (RowType) requestedSchema.toRowDataType().getLogicalType(); } else { return FlinkSchemaUtil.convert(schema); } } static IcebergStreamWriter createStreamWriter( SerializableSupplier
tableSupplier, FlinkWriteConf flinkWriteConf, RowType flinkRowType, List equalityFieldIds) { Preconditions.checkArgument(tableSupplier != null, "Iceberg table supplier shouldn't be null"); Table initTable = tableSupplier.get(); FileFormat format = flinkWriteConf.dataFileFormat(); TaskWriterFactory taskWriterFactory = new RowDataTaskWriterFactory( tableSupplier, flinkRowType, flinkWriteConf.targetDataFileSize(), format, writeProperties(initTable, format, flinkWriteConf), equalityFieldIds, flinkWriteConf.upsertMode()); return new IcebergStreamWriter<>(initTable.name(), taskWriterFactory); } /** * Based on the {@link FileFormat} overwrites the table level compression properties for the table * write. * * @param table The table to get the table level settings * @param format The FileFormat to use * @param conf The write configuration * @return The properties to use for writing */ private static Map writeProperties( Table table, FileFormat format, FlinkWriteConf conf) { Map writeProperties = Maps.newHashMap(table.properties()); switch (format) { case PARQUET: writeProperties.put(PARQUET_COMPRESSION, conf.parquetCompressionCodec()); String parquetCompressionLevel = conf.parquetCompressionLevel(); if (parquetCompressionLevel != null) { writeProperties.put(PARQUET_COMPRESSION_LEVEL, parquetCompressionLevel); } break; case AVRO: writeProperties.put(AVRO_COMPRESSION, conf.avroCompressionCodec()); String avroCompressionLevel = conf.avroCompressionLevel(); if (avroCompressionLevel != null) { writeProperties.put(AVRO_COMPRESSION_LEVEL, conf.avroCompressionLevel()); } break; case ORC: writeProperties.put(ORC_COMPRESSION, conf.orcCompressionCodec()); writeProperties.put(ORC_COMPRESSION_STRATEGY, conf.orcCompressionStrategy()); break; default: throw new IllegalArgumentException(String.format("Unknown file format %s", format)); } return writeProperties; } }