001/*
002 * Licensed to the Apache Software Foundation (ASF) under one
003 * or more contributor license agreements.  See the NOTICE file
004 * distributed with this work for additional information
005 * regarding copyright ownership.  The ASF licenses this file
006 * to you under the Apache License, Version 2.0 (the
007 * "License"); you may not use this file except in compliance
008 * with the License.  You may obtain a copy of the License at
009 *
010 *     http://www.apache.org/licenses/LICENSE-2.0
011 *
012 * Unless required by applicable law or agreed to in writing, software
013 * distributed under the License is distributed on an "AS IS" BASIS,
014 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
015 * See the License for the specific language governing permissions and
016 * limitations under the License.
017 */
018package org.apache.hadoop.hbase.regionserver.storefiletracker;
019
020import java.io.IOException;
021import java.util.Collection;
022import java.util.List;
023import org.apache.hadoop.fs.FileStatus;
024import org.apache.hadoop.fs.Path;
025import org.apache.hadoop.hbase.client.TableDescriptorBuilder;
026import org.apache.hadoop.hbase.io.Reference;
027import org.apache.hadoop.hbase.regionserver.CreateStoreFileWriterParams;
028import org.apache.hadoop.hbase.regionserver.StoreFileInfo;
029import org.apache.hadoop.hbase.regionserver.StoreFileWriter;
030import org.apache.yetus.audience.InterfaceAudience;
031
032/**
033 * An interface to define how we track the store files for a give store.
034 * <p/>
035 * In the old time, we will write store to a tmp directory first, and then rename it to the actual
036 * data file. And once a store file is under data directory, we will consider it as 'committed'. And
037 * we need to do listing when loading store files.
038 * <p/>
039 * When cloud age is coming, now we want to store the store files on object storage, where rename
040 * and list are not as cheap as on HDFS, especially rename. Although introducing a metadata
041 * management layer for object storage could solve the problem, but we still want HBase to run on
042 * pure object storage, so here we introduce this interface to abstract how we track the store
043 * files. For the old implementation, we just persist nothing here, and do listing to load store
044 * files. When running on object storage, we could persist the store file list in a system region,
045 * or in a file on the object storage, to make it possible to write directly into the data directory
046 * to avoid renaming, and also avoid listing when loading store files.
047 * <p/>
048 * The implementation requires to be thread safe as flush and compaction may occur as the same time,
049 * and we could also do multiple compactions at the same time. As the implementation may choose to
050 * persist the store file list to external storage, which could be slow, it is the duty for the
051 * callers to not call it inside a lock which may block normal read/write requests.
052 */
053@InterfaceAudience.Private
054public interface StoreFileTracker {
055  /**
056   * Load the store files list when opening a region.
057   */
058  List<StoreFileInfo> load() throws IOException;
059
060  /**
061   * Add new store files.
062   * <p/>
063   * Used for flush and bulk load.
064   */
065  void add(Collection<StoreFileInfo> newFiles) throws IOException;
066
067  /**
068   * Add new store files and remove compacted store files after compaction.
069   */
070  void replace(Collection<StoreFileInfo> compactedFiles, Collection<StoreFileInfo> newFiles)
071    throws IOException;
072
073  /**
074   * Set the store files.
075   */
076  void set(List<StoreFileInfo> files) throws IOException;
077
078  /**
079   * Create a writer for writing new store files.
080   * @return Writer for a new StoreFile
081   */
082  StoreFileWriter createWriter(CreateStoreFileWriterParams params) throws IOException;
083
084  /**
085   * Adds StoreFileTracker implementations specific configurations into the table descriptor.
086   * <p/>
087   * This is used to avoid accidentally data loss when changing the cluster level store file tracker
088   * implementation, and also possible misconfiguration between master and region servers.
089   * <p/>
090   * See HBASE-26246 for more details.
091   * @param builder The table descriptor builder for the given table.
092   */
093  TableDescriptorBuilder updateWithTrackerConfigs(TableDescriptorBuilder builder);
094
095  /**
096   * Whether the implementation of this tracker requires you to write to temp directory first, i.e,
097   * does not allow broken store files under the actual data directory.
098   */
099  boolean requireWritingToTmpDirFirst();
100
101  Reference createReference(Reference reference, Path path) throws IOException;
102
103  /**
104   * Reads the reference file from the given path.
105   * @param path the {@link Path} to the reference file in the file system.
106   * @return a {@link Reference} that points at top/bottom half of a an hfile
107   */
108  Reference readReference(Path path) throws IOException;
109
110  /**
111   * Returns true if the specified family has reference files
112   * @return true if family contains reference files
113   */
114  boolean hasReferences() throws IOException;
115
116  StoreFileInfo getStoreFileInfo(final FileStatus fileStatus, final Path initialPath,
117    final boolean primaryReplica) throws IOException;
118
119  StoreFileInfo getStoreFileInfo(final Path initialPath, final boolean primaryReplica)
120    throws IOException;
121
122}