steveloughran commented on a change in pull request #24970: [SPARK-23977][SQL] Support High Performance S3A committers [test-hadoop3.2] URL: https://github.com/apache/spark/pull/24970#discussion_r310122570
########## File path: hadoop-cloud/src/hadoop-3/main/scala/org/apache/spark/internal/io/cloud/PathOutputCommitProtocol.scala ########## @@ -0,0 +1,263 @@ +/* + * 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.spark.internal.io.cloud + +import java.io.IOException + +import org.apache.hadoop.fs.Path +import org.apache.hadoop.mapreduce.{JobContext, TaskAttemptContext} +import org.apache.hadoop.mapreduce.lib.output.{FileOutputCommitter, PathOutputCommitter, PathOutputCommitterFactory} + +import org.apache.spark.internal.io.{FileCommitProtocol, HadoopMapReduceCommitProtocol} + +/** + * Spark Commit protocol for Path Output Committers. + * This committer will work with the `FileOutputCommitter` and subclasses. + * All implementations *must* be serializable. + * + * Rather than ask the `FileOutputFormat` for a committer, it uses the + * `org.apache.hadoop.mapreduce.lib.output.PathOutputCommitterFactory` factory + * API to create the committer. + * + * In `setupCommitter` the factory is identified and instantiated to; + * this factory then creates the actual committer implementation. + * + * @constructor Instantiate. dynamic partition overwrite is not supported, + * so that committers for stores which do not support rename + * will not get confused. + * @param jobId job + * @param dest destination + * @param dynamicPartitionOverwrite does the caller want support for dynamic + * partition overwrite. If so, it will be + * refused. + * @throws IOException when an unsupported dynamicPartitionOverwrite option is supplied. + */ +class PathOutputCommitProtocol( + jobId: String, + dest: String, + dynamicPartitionOverwrite: Boolean = false) + extends HadoopMapReduceCommitProtocol(jobId, dest, false) with Serializable { + + if (dynamicPartitionOverwrite) { + // until there's explicit extensions to the PathOutputCommitProtocols + // to support the spark mechanism, it's left to the individual committer + // choice to handle partitioning. + throw new IOException(PathOutputCommitProtocol.UNSUPPORTED) + } + + /** The committer created. */ + @transient private var committer: PathOutputCommitter = _ + + require(dest != null, "Null destination specified") + + private[cloud] val destination: String = dest + + /** The destination path. This is serializable in Hadoop 3. */ + private[cloud] val destPath: Path = new Path(destination) + + logDebug(s"Instantiated committer with job ID=$jobId;" + + s" destination=$destPath;" + + s" dynamicPartitionOverwrite=$dynamicPartitionOverwrite") + + import PathOutputCommitProtocol._ + + /** + * Set up the committer. + * This creates it by talking directly to the Hadoop factories, instead + * of the V1 `mapred.FileOutputFormat` methods. + * @param context task attempt + * @return the committer to use. This will always be a subclass of + * [[PathOutputCommitter]]. + */ + override protected def setupCommitter(context: TaskAttemptContext): PathOutputCommitter = { + Review comment: done ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
