1 package org.apache.lucene.benchmark.byTask.tasks;
4 * Licensed to the Apache Software Foundation (ASF) under one or more
5 * contributor license agreements. See the NOTICE file distributed with
6 * this work for additional information regarding copyright ownership.
7 * The ASF licenses this file to You under the Apache License, Version 2.0
8 * (the "License"); you may not use this file except in compliance with
9 * the License. You may obtain a copy of the License at
11 * http://www.apache.org/licenses/LICENSE-2.0
13 * Unless required by applicable law or agreed to in writing, software
14 * distributed under the License is distributed on an "AS IS" BASIS,
15 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
16 * See the License for the specific language governing permissions and
17 * limitations under the License.
20 import org.apache.lucene.benchmark.byTask.PerfRunData;
21 import org.apache.lucene.benchmark.byTask.utils.Config;
22 import org.apache.lucene.index.CorruptIndexException;
23 import org.apache.lucene.index.IndexCommit;
24 import org.apache.lucene.index.IndexDeletionPolicy;
25 import org.apache.lucene.index.IndexWriter;
26 import org.apache.lucene.index.IndexWriterConfig;
27 import org.apache.lucene.index.LogMergePolicy;
28 import org.apache.lucene.index.MergeScheduler;
29 import org.apache.lucene.index.ConcurrentMergeScheduler;
30 import org.apache.lucene.index.MergePolicy;
31 import org.apache.lucene.index.NoDeletionPolicy;
32 import org.apache.lucene.index.NoMergePolicy;
33 import org.apache.lucene.index.NoMergeScheduler;
34 import org.apache.lucene.index.IndexWriterConfig.OpenMode;
35 import org.apache.lucene.index.TieredMergePolicy;
36 import org.apache.lucene.store.LockObtainFailedException;
37 import org.apache.lucene.util.Version;
39 import java.io.BufferedOutputStream;
41 import java.io.FileOutputStream;
42 import java.io.IOException;
43 import java.io.PrintStream;
46 * Create an index. <br>
47 * Other side effects: index writer object in perfRunData is set. <br>
48 * Relevant properties: <code>merge.factor (default 10),
49 * max.buffered (default no flush), max.field.length (default
50 * 10,000 tokens), max.field.length, compound (default true), ram.flush.mb [default 0],
51 * merge.policy (default org.apache.lucene.index.LogByteSizeMergePolicy),
52 * merge.scheduler (default
53 * org.apache.lucene.index.ConcurrentMergeScheduler),
54 * concurrent.merge.scheduler.max.thread.count and
55 * concurrent.merge.scheduler.max.merge.count (defaults per
56 * ConcurrentMergeScheduler) </code>.
58 * This task also supports a "writer.info.stream" property with the following
61 * <li>SystemOut - sets {@link IndexWriter#setInfoStream(java.io.PrintStream)}
62 * to {@link System#out}.
63 * <li>SystemErr - sets {@link IndexWriter#setInfoStream(java.io.PrintStream)}
64 * to {@link System#err}.
65 * <li><file_name> - attempts to create a file given that name and sets
66 * {@link IndexWriter#setInfoStream(java.io.PrintStream)} to that file. If this
67 * denotes an invalid file name, or some error occurs, an exception will be
71 public class CreateIndexTask extends PerfTask {
73 public CreateIndexTask(PerfRunData runData) {
77 public static IndexDeletionPolicy getIndexDeletionPolicy(Config config) {
78 String deletionPolicyName = config.get("deletion.policy", "org.apache.lucene.index.KeepOnlyLastCommitDeletionPolicy");
79 if (deletionPolicyName.equals(NoDeletionPolicy.class.getName())) {
80 return NoDeletionPolicy.INSTANCE;
83 return Class.forName(deletionPolicyName).asSubclass(IndexDeletionPolicy.class).newInstance();
84 } catch (Exception e) {
85 throw new RuntimeException("unable to instantiate class '" + deletionPolicyName + "' as IndexDeletionPolicy", e);
91 public int doLogic() throws IOException {
92 PerfRunData runData = getRunData();
93 Config config = runData.getConfig();
94 runData.setIndexWriter(configureWriter(config, runData, OpenMode.CREATE, null));
98 public static IndexWriterConfig createWriterConfig(Config config, PerfRunData runData, OpenMode mode, IndexCommit commit) {
99 Version version = Version.valueOf(config.get("writer.version", Version.LUCENE_31.toString()));
100 IndexWriterConfig iwConf = new IndexWriterConfig(version, runData.getAnalyzer());
101 iwConf.setOpenMode(mode);
102 IndexDeletionPolicy indexDeletionPolicy = getIndexDeletionPolicy(config);
103 iwConf.setIndexDeletionPolicy(indexDeletionPolicy);
105 iwConf.setIndexCommit(commit);
108 final String mergeScheduler = config.get("merge.scheduler",
109 "org.apache.lucene.index.ConcurrentMergeScheduler");
110 if (mergeScheduler.equals(NoMergeScheduler.class.getName())) {
111 iwConf.setMergeScheduler(NoMergeScheduler.INSTANCE);
114 iwConf.setMergeScheduler(Class.forName(mergeScheduler).asSubclass(MergeScheduler.class).newInstance());
115 } catch (Exception e) {
116 throw new RuntimeException("unable to instantiate class '" + mergeScheduler + "' as merge scheduler", e);
119 if (mergeScheduler.equals("org.apache.lucene.index.ConcurrentMergeScheduler")) {
120 ConcurrentMergeScheduler cms = (ConcurrentMergeScheduler) iwConf.getMergeScheduler();
121 int v = config.get("concurrent.merge.scheduler.max.thread.count", -1);
123 cms.setMaxThreadCount(v);
125 v = config.get("concurrent.merge.scheduler.max.merge.count", -1);
127 cms.setMaxMergeCount(v);
132 final String mergePolicy = config.get("merge.policy",
133 "org.apache.lucene.index.LogByteSizeMergePolicy");
134 boolean isCompound = config.get("compound", true);
135 if (mergePolicy.equals(NoMergePolicy.class.getName())) {
136 iwConf.setMergePolicy(isCompound ? NoMergePolicy.COMPOUND_FILES : NoMergePolicy.NO_COMPOUND_FILES);
139 iwConf.setMergePolicy(Class.forName(mergePolicy).asSubclass(MergePolicy.class).newInstance());
140 } catch (Exception e) {
141 throw new RuntimeException("unable to instantiate class '" + mergePolicy + "' as merge policy", e);
143 if(iwConf.getMergePolicy() instanceof LogMergePolicy) {
144 LogMergePolicy logMergePolicy = (LogMergePolicy) iwConf.getMergePolicy();
145 logMergePolicy.setUseCompoundFile(isCompound);
146 logMergePolicy.setMergeFactor(config.get("merge.factor",OpenIndexTask.DEFAULT_MERGE_PFACTOR));
147 } else if(iwConf.getMergePolicy() instanceof TieredMergePolicy) {
148 TieredMergePolicy tieredMergePolicy = (TieredMergePolicy) iwConf.getMergePolicy();
149 tieredMergePolicy.setUseCompoundFile(isCompound);
152 final double ramBuffer = config.get("ram.flush.mb",OpenIndexTask.DEFAULT_RAM_FLUSH_MB);
153 final int maxBuffered = config.get("max.buffered",OpenIndexTask.DEFAULT_MAX_BUFFERED);
154 if (maxBuffered == IndexWriterConfig.DISABLE_AUTO_FLUSH) {
155 iwConf.setRAMBufferSizeMB(ramBuffer);
156 iwConf.setMaxBufferedDocs(maxBuffered);
158 iwConf.setMaxBufferedDocs(maxBuffered);
159 iwConf.setRAMBufferSizeMB(ramBuffer);
165 public static IndexWriter configureWriter(Config config, PerfRunData runData, OpenMode mode, IndexCommit commit) throws CorruptIndexException, LockObtainFailedException, IOException {
166 IndexWriter writer = new IndexWriter(runData.getDirectory(), createWriterConfig(config, runData, mode, commit));
167 writer.setMaxFieldLength(config.get("max.field.length", OpenIndexTask.DEFAULT_MAX_FIELD_LENGTH));
169 String infoStreamVal = config.get("writer.info.stream", null);
170 if (infoStreamVal != null) {
171 if (infoStreamVal.equals("SystemOut")) {
172 writer.setInfoStream(System.out);
173 } else if (infoStreamVal.equals("SystemErr")) {
174 writer.setInfoStream(System.err);
176 File f = new File(infoStreamVal).getAbsoluteFile();
177 writer.setInfoStream(new PrintStream(new BufferedOutputStream(new FileOutputStream(f))));