View Javadoc

1   /*
2    * Licensed to the Apache Software Foundation (ASF) under one or more
3    * contributor license agreements. See the NOTICE file distributed with this
4    * work for additional information regarding copyright ownership. The ASF
5    * licenses this file to you under the Apache License, Version 2.0 (the
6    * "License"); you may not use this file except in compliance with the License.
7    * You may obtain a copy of the License at
8    *
9    * http://www.apache.org/licenses/LICENSE-2.0
10   *
11   * Unless required by applicable law or agreed to in writing, software
12   * distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
13   * WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
14   * License for the specific language governing permissions and limitations
15   * under the License.
16   */
17  package org.apache.hadoop.hbase.util.test;
18  
19  import java.nio.ByteBuffer;
20  import java.util.ArrayList;
21  import java.util.Collections;
22  import java.util.HashMap;
23  import java.util.List;
24  import java.util.Map;
25  import java.util.Random;
26  
27  import org.apache.hadoop.classification.InterfaceAudience;
28  import org.apache.hadoop.hbase.KeyValue;
29  import org.apache.hadoop.hbase.Tag;
30  import org.apache.hadoop.hbase.util.ByteBufferUtils;
31  import org.apache.hadoop.io.WritableUtils;
32  
33  import com.google.common.primitives.Bytes;
34  
35  /**
36   * Generate list of key values which are very useful to test data block encoding
37   * and compression.
38   */
39  @edu.umd.cs.findbugs.annotations.SuppressWarnings(
40      value="RV_ABSOLUTE_VALUE_OF_RANDOM_INT",
41      justification="Should probably fix")
42  @InterfaceAudience.Private
43  public class RedundantKVGenerator {
44    // row settings
45    static byte[] DEFAULT_COMMON_PREFIX = new byte[0];
46    static int DEFAULT_NUMBER_OF_ROW_PREFIXES = 10;
47    static int DEFAULT_AVERAGE_PREFIX_LENGTH = 6;
48    static int DEFAULT_PREFIX_LENGTH_VARIANCE = 3;
49    static int DEFAULT_AVERAGE_SUFFIX_LENGTH = 3;
50    static int DEFAULT_SUFFIX_LENGTH_VARIANCE = 3;
51    static int DEFAULT_NUMBER_OF_ROW = 500;
52  
53    // qualifier
54    static float DEFAULT_CHANCE_FOR_SAME_QUALIFIER = 0.5f;
55    static float DEFAULT_CHANCE_FOR_SIMILIAR_QUALIFIER = 0.4f;
56    static int DEFAULT_AVERAGE_QUALIFIER_LENGTH = 9;
57    static int DEFAULT_QUALIFIER_LENGTH_VARIANCE = 3;
58  
59    static int DEFAULT_COLUMN_FAMILY_LENGTH = 9;
60    static int DEFAULT_VALUE_LENGTH = 8;
61    static float DEFAULT_CHANCE_FOR_ZERO_VALUE = 0.5f;
62  
63    static int DEFAULT_BASE_TIMESTAMP_DIVIDE = 1000000;
64    static int DEFAULT_TIMESTAMP_DIFF_SIZE = 100000000;
65  
66    /**
67     * Default constructor, assumes all parameters from class constants.
68     */
69    public RedundantKVGenerator() {
70      this(new Random(42L),
71          DEFAULT_NUMBER_OF_ROW_PREFIXES,
72          DEFAULT_AVERAGE_PREFIX_LENGTH,
73          DEFAULT_PREFIX_LENGTH_VARIANCE,
74          DEFAULT_AVERAGE_SUFFIX_LENGTH,
75          DEFAULT_SUFFIX_LENGTH_VARIANCE,
76          DEFAULT_NUMBER_OF_ROW,
77  
78          DEFAULT_CHANCE_FOR_SAME_QUALIFIER,
79          DEFAULT_CHANCE_FOR_SIMILIAR_QUALIFIER,
80          DEFAULT_AVERAGE_QUALIFIER_LENGTH,
81          DEFAULT_QUALIFIER_LENGTH_VARIANCE,
82  
83          DEFAULT_COLUMN_FAMILY_LENGTH,
84          DEFAULT_VALUE_LENGTH,
85          DEFAULT_CHANCE_FOR_ZERO_VALUE,
86  
87          DEFAULT_BASE_TIMESTAMP_DIVIDE,
88          DEFAULT_TIMESTAMP_DIFF_SIZE
89      );
90    }
91  
92  
93    /**
94     * Various configuration options for generating key values
95     * @param randomizer pick things by random
96     */
97    public RedundantKVGenerator(Random randomizer,
98        int numberOfRowPrefixes,
99        int averagePrefixLength,
100       int prefixLengthVariance,
101       int averageSuffixLength,
102       int suffixLengthVariance,
103       int numberOfRows,
104 
105       float chanceForSameQualifier,
106       float chanceForSimiliarQualifier,
107       int averageQualifierLength,
108       int qualifierLengthVariance,
109 
110       int columnFamilyLength,
111       int valueLength,
112       float chanceForZeroValue,
113 
114       int baseTimestampDivide,
115       int timestampDiffSize
116       ) {
117     this.randomizer = randomizer;
118 
119     this.commonPrefix = DEFAULT_COMMON_PREFIX;
120     this.numberOfRowPrefixes = numberOfRowPrefixes;
121     this.averagePrefixLength = averagePrefixLength;
122     this.prefixLengthVariance = prefixLengthVariance;
123     this.averageSuffixLength = averageSuffixLength;
124     this.suffixLengthVariance = suffixLengthVariance;
125     this.numberOfRows = numberOfRows;
126 
127     this.chanceForSameQualifier = chanceForSameQualifier;
128     this.chanceForSimilarQualifier = chanceForSimiliarQualifier;
129     this.averageQualifierLength = averageQualifierLength;
130     this.qualifierLengthVariance = qualifierLengthVariance;
131 
132     this.columnFamilyLength = columnFamilyLength;
133     this.valueLength = valueLength;
134     this.chanceForZeroValue = chanceForZeroValue;
135 
136     this.baseTimestampDivide = baseTimestampDivide;
137     this.timestampDiffSize = timestampDiffSize;
138   }
139 
140   /** Used to generate dataset */
141   private Random randomizer;
142 
143   // row settings
144   private byte[] commonPrefix;//global prefix before rowPrefixes
145   private int numberOfRowPrefixes;
146   private int averagePrefixLength = 6;
147   private int prefixLengthVariance = 3;
148   private int averageSuffixLength = 3;
149   private int suffixLengthVariance = 3;
150   private int numberOfRows = 500;
151 
152   //family
153   private byte[] family;
154 
155   // qualifier
156   private float chanceForSameQualifier = 0.5f;
157   private float chanceForSimilarQualifier = 0.4f;
158   private int averageQualifierLength = 9;
159   private int qualifierLengthVariance = 3;
160 
161   private int columnFamilyLength = 9;
162   private int valueLength = 8;
163   private float chanceForZeroValue = 0.5f;
164 
165   private int baseTimestampDivide = 1000000;
166   private int timestampDiffSize = 100000000;
167 
168   private List<byte[]> generateRows() {
169     // generate prefixes
170     List<byte[]> prefixes = new ArrayList<byte[]>();
171     prefixes.add(new byte[0]);
172     for (int i = 1; i < numberOfRowPrefixes; ++i) {
173       int prefixLength = averagePrefixLength;
174       prefixLength += randomizer.nextInt(2 * prefixLengthVariance + 1) -
175           prefixLengthVariance;
176       byte[] newPrefix = new byte[prefixLength];
177       randomizer.nextBytes(newPrefix);
178       byte[] newPrefixWithCommon = newPrefix;
179       prefixes.add(newPrefixWithCommon);
180     }
181 
182     // generate rest of the row
183     List<byte[]> rows = new ArrayList<byte[]>();
184     for (int i = 0; i < numberOfRows; ++i) {
185       int suffixLength = averageSuffixLength;
186       suffixLength += randomizer.nextInt(2 * suffixLengthVariance + 1) -
187           suffixLengthVariance;
188       int randomPrefix = randomizer.nextInt(prefixes.size());
189       byte[] row = new byte[prefixes.get(randomPrefix).length +
190                             suffixLength];
191       byte[] rowWithCommonPrefix = Bytes.concat(commonPrefix, row);
192       rows.add(rowWithCommonPrefix);
193     }
194 
195     return rows;
196   }
197 
198   /**
199    * Generate test data useful to test encoders.
200    * @param howMany How many Key values should be generated.
201    * @return sorted list of key values
202    */
203   public List<KeyValue> generateTestKeyValues(int howMany) {
204     return generateTestKeyValues(howMany, false);
205   }
206   /**
207    * Generate test data useful to test encoders.
208    * @param howMany How many Key values should be generated.
209    * @return sorted list of key values
210    */
211   public List<KeyValue> generateTestKeyValues(int howMany, boolean useTags) {
212     List<KeyValue> result = new ArrayList<KeyValue>();
213 
214     List<byte[]> rows = generateRows();
215     Map<Integer, List<byte[]>> rowsToQualifier = new HashMap<Integer, List<byte[]>>();
216 
217     if(family==null){
218       family = new byte[columnFamilyLength];
219       randomizer.nextBytes(family);
220     }
221 
222     long baseTimestamp = Math.abs(randomizer.nextInt()) / baseTimestampDivide;
223 
224     byte[] value = new byte[valueLength];
225 
226     for (int i = 0; i < howMany; ++i) {
227       long timestamp = baseTimestamp;
228       if(timestampDiffSize > 0){
229         timestamp += randomizer.nextInt(timestampDiffSize);
230       }
231       Integer rowId = randomizer.nextInt(rows.size());
232       byte[] row = rows.get(rowId);
233 
234       // generate qualifier, sometimes it is same, sometimes similar,
235       // occasionally completely different
236       byte[] qualifier;
237       float qualifierChance = randomizer.nextFloat();
238       if (!rowsToQualifier.containsKey(rowId)
239           || qualifierChance > chanceForSameQualifier + chanceForSimilarQualifier) {
240         int qualifierLength = averageQualifierLength;
241         qualifierLength += randomizer.nextInt(2 * qualifierLengthVariance + 1)
242             - qualifierLengthVariance;
243         qualifier = new byte[qualifierLength];
244         randomizer.nextBytes(qualifier);
245 
246         // add it to map
247         if (!rowsToQualifier.containsKey(rowId)) {
248           rowsToQualifier.put(rowId, new ArrayList<byte[]>());
249         }
250         rowsToQualifier.get(rowId).add(qualifier);
251       } else if (qualifierChance > chanceForSameQualifier) {
252         // similar qualifier
253         List<byte[]> previousQualifiers = rowsToQualifier.get(rowId);
254         byte[] originalQualifier = previousQualifiers.get(randomizer.nextInt(previousQualifiers
255             .size()));
256 
257         qualifier = new byte[originalQualifier.length];
258         int commonPrefix = randomizer.nextInt(qualifier.length);
259         System.arraycopy(originalQualifier, 0, qualifier, 0, commonPrefix);
260         for (int j = commonPrefix; j < qualifier.length; ++j) {
261           qualifier[j] = (byte) (randomizer.nextInt() & 0xff);
262         }
263 
264         rowsToQualifier.get(rowId).add(qualifier);
265       } else {
266         // same qualifier
267         List<byte[]> previousQualifiers = rowsToQualifier.get(rowId);
268         qualifier = previousQualifiers.get(randomizer.nextInt(previousQualifiers.size()));
269       }
270 
271       if (randomizer.nextFloat() < chanceForZeroValue) {
272         for (int j = 0; j < value.length; ++j) {
273           value[j] = (byte) 0;
274         }
275       } else {
276         randomizer.nextBytes(value);
277       }
278 
279       if (useTags) {
280         result.add(new KeyValue(row, family, qualifier, timestamp, value, new Tag[] { new Tag(
281             (byte) 1, "value1") }));
282       } else {
283         result.add(new KeyValue(row, family, qualifier, timestamp, value));
284       }
285     }
286 
287     Collections.sort(result, KeyValue.COMPARATOR);
288 
289     return result;
290   }
291 
292   /**
293    * Convert list of KeyValues to byte buffer.
294    * @param keyValues list of KeyValues to be converted.
295    * @return buffer with content from key values
296    */
297   public static ByteBuffer convertKvToByteBuffer(List<KeyValue> keyValues,
298       boolean includesMemstoreTS) {
299     int totalSize = 0;
300     for (KeyValue kv : keyValues) {
301       totalSize += kv.getLength();
302       if (includesMemstoreTS) {
303         totalSize += WritableUtils.getVIntSize(kv.getMvccVersion());
304       }
305     }
306 
307     ByteBuffer result = ByteBuffer.allocate(totalSize);
308     for (KeyValue kv : keyValues) {
309       result.put(kv.getBuffer(), kv.getOffset(), kv.getLength());
310       if (includesMemstoreTS) {
311         ByteBufferUtils.writeVLong(result, kv.getMvccVersion());
312       }
313     }
314     return result;
315   }
316   
317   
318   /************************ get/set ***********************************/
319   
320   public RedundantKVGenerator setCommonPrefix(byte[] prefix){
321     this.commonPrefix = prefix;
322     return this;
323   }
324 
325   public RedundantKVGenerator setRandomizer(Random randomizer) {
326     this.randomizer = randomizer;
327     return this;
328   }
329 
330   public RedundantKVGenerator setNumberOfRowPrefixes(int numberOfRowPrefixes) {
331     this.numberOfRowPrefixes = numberOfRowPrefixes;
332     return this;
333   }
334 
335   public RedundantKVGenerator setAveragePrefixLength(int averagePrefixLength) {
336     this.averagePrefixLength = averagePrefixLength;
337     return this;
338   }
339 
340   public RedundantKVGenerator setPrefixLengthVariance(int prefixLengthVariance) {
341     this.prefixLengthVariance = prefixLengthVariance;
342     return this;
343   }
344 
345   public RedundantKVGenerator setAverageSuffixLength(int averageSuffixLength) {
346     this.averageSuffixLength = averageSuffixLength;
347     return this;
348   }
349 
350   public RedundantKVGenerator setSuffixLengthVariance(int suffixLengthVariance) {
351     this.suffixLengthVariance = suffixLengthVariance;
352     return this;
353   }
354 
355   public RedundantKVGenerator setNumberOfRows(int numberOfRows) {
356     this.numberOfRows = numberOfRows;
357     return this;
358   }
359 
360   public RedundantKVGenerator setChanceForSameQualifier(float chanceForSameQualifier) {
361     this.chanceForSameQualifier = chanceForSameQualifier;
362     return this;
363   }
364 
365   public RedundantKVGenerator setChanceForSimilarQualifier(float chanceForSimiliarQualifier) {
366     this.chanceForSimilarQualifier = chanceForSimiliarQualifier;
367     return this;
368   }
369 
370   public RedundantKVGenerator setAverageQualifierLength(int averageQualifierLength) {
371     this.averageQualifierLength = averageQualifierLength;
372     return this;
373   }
374 
375   public RedundantKVGenerator setQualifierLengthVariance(int qualifierLengthVariance) {
376     this.qualifierLengthVariance = qualifierLengthVariance;
377     return this;
378   }
379 
380   public RedundantKVGenerator setColumnFamilyLength(int columnFamilyLength) {
381     this.columnFamilyLength = columnFamilyLength;
382     return this;
383   }
384 
385   public RedundantKVGenerator setFamily(byte[] family) {
386     this.family = family;
387     this.columnFamilyLength = family.length;
388     return this;
389   }
390 
391   public RedundantKVGenerator setValueLength(int valueLength) {
392     this.valueLength = valueLength;
393     return this;
394   }
395 
396   public RedundantKVGenerator setChanceForZeroValue(float chanceForZeroValue) {
397     this.chanceForZeroValue = chanceForZeroValue;
398     return this;
399   }
400 
401   public RedundantKVGenerator setBaseTimestampDivide(int baseTimestampDivide) {
402     this.baseTimestampDivide = baseTimestampDivide;
403     return this;
404   }
405 
406   public RedundantKVGenerator setTimestampDiffSize(int timestampDiffSize) {
407     this.timestampDiffSize = timestampDiffSize;
408     return this;
409   }
410   
411 }