amogh-jahagirdar commented on code in PR #15448: URL: https://github.com/apache/iceberg/pull/15448#discussion_r2954691226
########## spark/v4.1/spark/src/main/java/org/apache/iceberg/spark/source/SerializableFileIOWithSize.java: ########## @@ -0,0 +1,120 @@ +/* + * 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.source; + +import java.util.Map; +import java.util.function.Function; +import org.apache.hadoop.conf.Configuration; +import org.apache.iceberg.hadoop.HadoopConfigurable; +import org.apache.iceberg.io.FileIO; +import org.apache.iceberg.io.InputFile; +import org.apache.iceberg.io.OutputFile; +import org.apache.iceberg.util.SerializableSupplier; +import org.apache.spark.util.KnownSizeEstimation; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +/** + * This class provides a serializable {@link FileIO} with a known size estimate. Spark calls its + * {@link org.apache.spark.util.SizeEstimator} class when broadcasting variables and this can be an + * expensive operation, so providing a known size estimate allows that operation to be skipped. + * + * <p>This class also implements {@link AutoCloseable} to avoid leaking resources upon broadcasting. + * Broadcast variables are destroyed and cleaned up on the driver and executors once they are + * garbage collected on the driver. The implementation ensures only resources used by copies of the + * main {@link FileIO} are released. + */ +class SerializableFileIOWithSize + implements FileIO, HadoopConfigurable, KnownSizeEstimation, AutoCloseable { + private static final Logger LOG = LoggerFactory.getLogger(SerializableFileIOWithSize.class); + private static final long SIZE_ESTIMATE = 32_768L; + private final transient Object serializationMarker; + private final FileIO fileIO; + + private SerializableFileIOWithSize(FileIO fileIO) { + this.fileIO = fileIO; + this.serializationMarker = new Object(); + } + + @Override + public long estimatedSize() { + return SIZE_ESTIMATE; + } + + public static FileIO wrap(FileIO fileIO) { + return new SerializableFileIOWithSize(fileIO); + } + + @Override + public void close() { + if (null == serializationMarker) { + LOG.info("Closing FileIO"); Review Comment: Minor: Not sure we really need the log ########## spark/v4.1/spark-extensions/src/test/java/org/apache/iceberg/spark/source/TestRemoteScanPlanning.java: ########## @@ -38,8 +49,6 @@ protected static Object[][] parameters() { ImmutableMap.builder() .putAll(SparkCatalogConfig.REST.properties()) .put(CatalogProperties.URI, restCatalog.properties().get(CatalogProperties.URI)) - // this flag is typically only set by the server, but we set it from the client for Review Comment: So do we still need the explicit client setup below then? Or do we need more machinery in the rest fixture to be able to create tables which will later kick back "scan-planning" required in the load table response? -- 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]
