shangxinli commented on code in PR #14435: URL: https://github.com/apache/iceberg/pull/14435#discussion_r2572003069
########## spark/v4.0/spark/src/main/java/org/apache/iceberg/spark/actions/SparkParquetFileMergeRunner.java: ########## @@ -0,0 +1,460 @@ +/* + * 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.spark.actions; + +import java.io.IOException; +import java.io.Serializable; +import java.nio.ByteBuffer; +import java.util.List; +import java.util.Map; +import java.util.Set; +import java.util.stream.Collectors; +import org.apache.hadoop.conf.Configuration; +import org.apache.iceberg.DataFile; +import org.apache.iceberg.DataFiles; +import org.apache.iceberg.FileFormat; +import org.apache.iceberg.MetadataColumns; +import org.apache.iceberg.Metrics; +import org.apache.iceberg.MetricsConfig; +import org.apache.iceberg.PartitionSpec; +import org.apache.iceberg.StructLike; +import org.apache.iceberg.Table; +import org.apache.iceberg.TableProperties; +import org.apache.iceberg.TableUtil; +import org.apache.iceberg.actions.RewriteFileGroup; +import org.apache.iceberg.encryption.EncryptedOutputFile; +import org.apache.iceberg.io.FileIO; +import org.apache.iceberg.io.InputFile; +import org.apache.iceberg.io.OutputFileFactory; +import org.apache.iceberg.parquet.ParquetFileMerger; +import org.apache.iceberg.parquet.ParquetUtil; +import org.apache.iceberg.relocated.com.google.common.annotations.VisibleForTesting; +import org.apache.iceberg.relocated.com.google.common.collect.Lists; +import org.apache.iceberg.relocated.com.google.common.collect.Sets; +import org.apache.iceberg.spark.FileRewriteCoordinator; +import org.apache.iceberg.types.Conversions; +import org.apache.iceberg.types.Types; +import org.apache.iceberg.util.PropertyUtil; +import org.apache.parquet.schema.MessageType; +import org.apache.spark.api.java.JavaRDD; +import org.apache.spark.api.java.JavaSparkContext; +import org.apache.spark.sql.SparkSession; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +/** + * Extension of SparkBinPackFileRewriteRunner that uses ParquetFileMerger for efficient row-group + * level merging of Parquet files when applicable. + * + * <p>This runner uses {@link ParquetFileMerger} to merge Parquet files at the row-group level + * without full deserialization, which is significantly faster than the standard Spark rewrite + * approach for Parquet files. + * + * <p>The decision to use this runner vs. SparkBinPackFileRewriteRunner is controlled by the + * configuration option {@code use-parquet-file-merger}. + */ +public class SparkParquetFileMergeRunner extends SparkBinPackFileRewriteRunner { + private static final Logger LOG = LoggerFactory.getLogger(SparkParquetFileMergeRunner.class); + private final FileRewriteCoordinator coordinator = FileRewriteCoordinator.get(); + + public SparkParquetFileMergeRunner(SparkSession spark, Table table) { + super(spark, table); + } + + @Override + public String description() { + return "PARQUET-MERGE"; + } + + @Override + protected void doRewrite(String groupId, RewriteFileGroup group) { + // Early validation: check if requirements are met + if (!canMerge(group)) { + LOG.info( + "Row-group merge requirements not met for group {}. Using standard Spark rewrite.", + groupId); + super.doRewrite(groupId, group); + return; + } + + // Requirements met - attempt row-group level merge + try { + LOG.info( + "Merging {} Parquet files using row-group level merge (group: {})", + group.rewrittenFiles().size(), + groupId); + mergeParquetFilesDistributed(groupId, group); + } catch (Exception e) { + LOG.warn( + "Row-group merge failed for group {}, falling back to standard Spark rewrite: {}", + groupId, + e.getMessage(), + e); + // Fallback to standard rewrite + super.doRewrite(groupId, group); + } + } + + /** + * Validates if the file group can use row-group level merging. + * + * <p>Requirements checked here: + * + * <ul> + * <li>Table must not have a sort order (no sorting or z-ordering) + * <li>Files must not have delete files or delete vectors + * </ul> + * + * <p>Additional requirements checked by ParquetFileMerger.canMerge: + * + * <ul> + * <li>All files must be valid Parquet format + * <li>Files must have compatible schemas + * <li>Files must not be encrypted + * <li>If a physical _row_id column exists, all values must be non-null + * </ul> + * + * @param group the file group to check + * @return true if row-group merging can be used, false otherwise + */ + @VisibleForTesting + boolean canMerge(RewriteFileGroup group) { + // Check if table has a sort order - row-group merge cannot preserve sort order + if (table().sortOrder().isSorted()) { + LOG.debug( + "Cannot use row-group merge: table has a sort order ({}). " + + "Row-group merging would not preserve the sort order.", + table().sortOrder()); + return false; + } + + // Check for delete files - row-group merge cannot apply deletes + boolean hasDeletes = group.fileScanTasks().stream().anyMatch(task -> !task.deletes().isEmpty()); + + if (hasDeletes) { + LOG.debug( + "Cannot use row-group merge: files have delete files or delete vectors. " + + "Row-group merging cannot apply deletes."); + return false; + } + + // Validate schema compatibility and other Parquet-specific requirements + try { + List<InputFile> inputFiles = + group.rewrittenFiles().stream() + .map(f -> table().io().newInputFile(f.path().toString())) + .collect(Collectors.toList()); + + // Validate files can be merged + boolean canMerge = ParquetFileMerger.canMerge(inputFiles); + + if (!canMerge) { + LOG.warn( + "Cannot use row-group merge for {} files. Falling back to standard rewrite. " + + "Reason: Parquet validation failed", + group.rewrittenFiles().size()); + } + + return canMerge; + } catch (Exception e) { + LOG.warn("Cannot use row-group merge: validation failed", e); + return false; + } + } + + /** + * Merges Parquet files in the group, respecting the expected output file count determined by the + * planner. Files are distributed evenly across the expected number of output files. + */ + private void mergeParquetFilesDistributed(String groupId, RewriteFileGroup group) { + PartitionSpec spec = table().specs().get(group.outputSpecId()); + StructLike partition = group.info().partition(); + int expectedOutputFiles = group.expectedOutputFiles(); + + LOG.info( + "Merging {} Parquet files into {} expected output files (group: {})", + group.rewrittenFiles().size(), + expectedOutputFiles, + groupId); + + // Check if table supports row lineage + boolean preserveRowLineage = TableUtil.supportsRowLineage(table()); + + // Distribute files evenly across expected output files (planner already determined the count) + List<List<DataFile>> fileBatches = + distributeFilesEvenly(group.rewrittenFiles(), expectedOutputFiles); + + // Get row group size from table properties + long rowGroupSize = + PropertyUtil.propertyAsLong( + table().properties(), + TableProperties.PARQUET_ROW_GROUP_SIZE_BYTES, + TableProperties.PARQUET_ROW_GROUP_SIZE_BYTES_DEFAULT); + + // Get column index truncate length from Hadoop Configuration (same as ParquetWriter) + Configuration hadoopConf = spark().sessionState().newHadoopConf(); + int columnIndexTruncateLength = + hadoopConf.getInt( + ParquetUtil.COLUMN_INDEX_TRUNCATE_LENGTH, + ParquetUtil.DEFAULT_COLUMN_INDEX_TRUNCATE_LENGTH); + + // Create merge tasks for each batch + List<MergeTaskInfo> mergeTasks = Lists.newArrayList(); + int batchIndex = 0; + for (List<DataFile> batch : fileBatches) { + String taskId = String.format("%s-%d", groupId, batchIndex); + + // Create unique OutputFileFactory for each task to avoid filename collisions + OutputFileFactory fileFactory = + OutputFileFactory.builderFor(table(), spec.specId(), batchIndex) + .format(FileFormat.PARQUET) + .build(); + + batchIndex++; + + List<String> filePaths = + batch.stream().map(f -> f.path().toString()).collect(Collectors.toList()); + + // Extract firstRowIds for row lineage preservation + List<Long> firstRowIds = null; + if (preserveRowLineage) { + firstRowIds = batch.stream().map(DataFile::firstRowId).collect(Collectors.toList()); + LOG.debug( + "Task {} will preserve row lineage with firstRowIds: {} (group: {})", + taskId, + firstRowIds, + groupId); + } + + mergeTasks.add( + new MergeTaskInfo( + filePaths, + rowGroupSize, + columnIndexTruncateLength, + firstRowIds, + fileFactory, + partition)); + } + + // Get FileIO for executors - table().io() is serializable + FileIO fileIO = table().io(); + + // Execute merges on executors in parallel + JavaSparkContext jsc = JavaSparkContext.fromSparkContext(spark().sparkContext()); + JavaRDD<MergeTaskInfo> taskRDD = jsc.parallelize(mergeTasks, mergeTasks.size()); + List<MergeResult> mergeResults = taskRDD.map(task -> mergeFilesForTask(task, fileIO)).collect(); Review Comment: This is a great point! There was a concurrent write risk during task retries. The OutputFileFactory was created on driver with hardcoded taskId=0 and serialized to executors. On retry, both old and new task executions had the same factory and same filename. I just moved OutputFileFactory creation to the executor and use TaskContext.taskAttemptId() which will generate a new id to be part of the file name. It is something like below. long taskAttemptId = TaskContext.get().taskAttemptId(); OutputFileFactory fileFactory = OutputFileFactory.builderFor(table, task.batchIndex(), taskAttemptId) .defaultSpec(task.spec()) .format(task.format()) .build(); -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
