galenwarren commented on a change in pull request #15599: URL: https://github.com/apache/flink/pull/15599#discussion_r619845422
########## File path: flink-filesystems/flink-gs-fs-hadoop/src/main/java/org/apache/flink/fs/gs/writer/GSRecoverableWriterState.java ########## @@ -0,0 +1,158 @@ +/* + * 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.flink.fs.gs.writer; + +import org.apache.flink.core.fs.RecoverableWriter; +import org.apache.flink.fs.gs.GSFileSystemOptions; +import org.apache.flink.util.Preconditions; + +import com.google.cloud.storage.BlobId; + +import java.util.ArrayList; +import java.util.List; +import java.util.UUID; +import java.util.stream.Collectors; + +/** The state of a recoverable write. */ +class GSRecoverableWriterState implements RecoverableWriter.ResumeRecoverable, Cloneable { + + /** The blob id to which the recoverable write operation is writing. */ + public final BlobId finalBlobId; + + /** The number of bytes that have been written so far. */ + public long bytesWritten; + + /** Indicates if the write has been closed. */ + public boolean closed; + + /** The object ids for the temporary objects that should be composed to form the final blob. */ + public final List<UUID> componentObjectIds; + + GSRecoverableWriterState( + BlobId finalBlobId, long bytesWritten, boolean closed, List<UUID> componentObjectIds) { + this.finalBlobId = Preconditions.checkNotNull(finalBlobId); + Preconditions.checkArgument(bytesWritten >= 0); + this.bytesWritten = bytesWritten; + this.closed = closed; + + // shallow copy the component object ids to ensure this state object exclusively + // manages the list of component object ids + this.componentObjectIds = new ArrayList<>(Preconditions.checkNotNull(componentObjectIds)); + } + + GSRecoverableWriterState(GSRecoverableWriterState state) { + this(state.finalBlobId, state.bytesWritten, state.closed, state.componentObjectIds); + } + + GSRecoverableWriterState(BlobId finalBlobId) { + this(finalBlobId, 0, false, new ArrayList<>()); + } + + /** + * Returns the temporary bucket name. If options specifies a temporary bucket name, we use that + * one; otherwise, we use the bucket name of the final blob. + * + * @param options The GS file system options + * @return The temporary bucket name + */ + String getTemporaryBucketName(GSFileSystemOptions options) { + return options.writerTemporaryBucketName.isEmpty() + ? finalBlobId.getBucket() + : options.writerTemporaryBucketName; + } Review comment: If a dedicated temp bucket is specified in the options, then, yes, the same temp bucket is used for all writers/states. But if no dedicated temp bucket is specified, then the bucket that is used for temp blobs is the same bucket to which the "final" blob is being written, so that would be different when different buckets are being written to by streaming file sinks, and it would depend on the finalBlobId that is part of the state. -- 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]
