Mapreduce执行机制之Map和Reduce

1、Mapper 类

 * Maps input key/value pairs to a set of intermediate key/value pairs.  
 * 
 * Maps are the individual tasks which transform input records into a 
 * intermediate records. The transformed intermediate records need not be of 
 * the same type as the input records. A given input pair may map to zero or 
 * many output pairs. 
 * 
 * The Hadoop Map-Reduce framework spawns one map task for each 
 * {@link InputSplit} generated by the {@link InputFormat} for the job.
 * Mapper implementations can access the {@link Configuration} for 
 * the job via the {@link JobContext#getConfiguration()}.
 * 
 * 
The framework first calls 
 * {@link #setup(org.apache.hadoop.mapreduce.Mapper.Context)}, followed by
 * {@link #map(Object, Object, org.apache.hadoop.mapreduce.Mapper.Context)}
 * for each key/value pair in the InputSplit. Finally 
 * {@link #cleanup(org.apache.hadoop.mapreduce.Mapper.Context)} is called.

 * 
 * All intermediate values associated with a given output key are 
 * subsequently grouped by the framework, and passed to a {@link Reducer} to  
 * determine the final output. Users can control the sorting and grouping by 
 * specifying two key {@link RawComparator} classes.
 *
 * The Mapper outputs are partitioned per 
 * Reducer. Users can control which keys (and hence records) go to 
 * which Reducer by implementing a custom {@link Partitioner}.
 * 
 * 
Users can optionally specify a combiner, via 
 * {@link Job#setCombinerClass(Class)}, to perform local aggregation of the 
 * intermediate outputs, which helps to cut down the amount of data transferred 
 * from the Mapper to the Reducer.
 * 
 * 
Applications can specify if and how the intermediate
 * outputs are to be compressed and which {@link CompressionCodec}s are to be
 * used via the Configuration.

 *  
 * If the job has zero
 * reduces then the output of the Mapper is directly written
 * to the {@link OutputFormat} without sorting by keys.

 * 
 * Example:
 * 
 * public class TokenCounterMapper 
 *     extends Mapper<Object, Text, Text, IntWritable>{
 *    
 *   private final static IntWritable one = new IntWritable(1);
 *   private Text word = new Text();
 *   
 *   public void map(Object key, Text value, Context context) throws IOException, InterruptedException {
 *     StringTokenizer itr = new StringTokenizer(value.toString());
 *     while (itr.hasMoreTokens()) {
 *       word.set(itr.nextToken());
 *       context.write(word, one);
 *     }
 *   }
 * }
 * 
 *
 * Applications may override the
 * {@link #run(org.apache.hadoop.mapreduce.Mapper.Context)} method to exert
 * greater control on map processing e.g. multi-threaded Mappers 
 * etc.




将输入键/值对映射到一组中间键/值对。 
   
 map是将输入记录转换为a  
中间记录。 转换后的中间记录不必是  
与输入记录的类型相同。 给定的输入对可以映射为零或  
 . txt / /输出>  
＊  
*  Hadoop map - reduce框架为每个映射生成一个映射任务  
* {@link InputFormat}为作业生成的{@link InputSplit}。  
* Mapper实现可以访问{@link Configuration}  
*该任务通过{@link JobContext#getConfiguration()}。  







**   mapper调用流程

* 
框架首先调用  
* {@link #setup(org.apache.hadoop.mapreduce.Mapper.Context)}，然后是  
* {@link #map(Object, Object, org.apache.hadoop.mapreduce.Mapper.Context)}  
*为InputSplit. value >中的每个键/值对。 最后  
* {@link #cleanup(org.apache.hadoop.mapreduce.Mapper.Context)}被调用  
＊  
* 
所有与给定输出键相关联的中间值是  
*随后被框架分组，并传递给{@link Reducer}  
*确定最终输出。 用户可以通过控制排序和分组  
 . *指定两个键{@link RawComparator}类  


** 1、Partioner分区

*  Mapper output are partitioned per . / 
 Mapper output  
* <代码>减速器> < /代码。 用户可以控制去哪个键(以及记录)  
Reducer通过实现一个自定义的{@link Partitioner}。  





*** 2、数据预合并Combiner

* 
用户可以选择指定一个合成器，通过  
* {@link Job#setCombinerClass(Class)}，执行局部聚合  
*中间输出，有助于减少传输的数据量  
从Mapper到Reducer. *  


** 3、压缩Compression
 
* 
应用程序可以指定是否以及如何使用中间体  
*输出将被压缩，哪个{@link CompressionCodec}将被压缩  
*通过Configuration.
 . Configuration

 
Mapper核心调用 
Reducer 类 
* Reduces a set of intermediate values which share a key to a smaller set of
 * values.  
 * 
 * Reducer implementations 
 * can access the {@link Configuration} for the job via the 
 * {@link JobContext#getConfiguration()} method.


 * Reducer has 3 primary phases:

 * 
 *   
 *   
 *   Shuffle
 *   
 *   The Reducer copies the sorted output from each 
 *   {@link Mapper} using HTTP across the network.

 *   

 *   
 *   
 *   Sort
 *   
 *   The framework merge sorts Reducer inputs by 
 *   keys 
 *   (since different Mappers may have output the same key).

 *   
 *   The shuffle and sort phases occur simultaneously i.e. while outputs are
 *   being fetched they are merged.
 *      
 *   SecondarySort
 *   
 *   To achieve a secondary sort on the values returned by the value 
 *   iterator, the application should extend the key with the secondary
 *   key and define a grouping comparator. The keys will be sorted using the
 *   entire key, but will be grouped using the grouping comparator to decide
 *   which keys and values are sent in the same call to reduce.The grouping 
 *   comparator is specified via 
 *   {@link Job#setGroupingComparatorClass(Class)}. The sort order is
 *   controlled by 
 *   {@link Job#setSortComparatorClass(Class)}.


*将一组共享密钥的中间值减少到更小的一组  
*值。  
＊  
* < p > <代码>减速器> < /代码的实现  
. *可以访问任务的{@link配置}  
* {@link JobContext#getConfiguration()}方法  
 
Reducer有3个主要阶段:
  
* < ol >  
李* < >  
＊  
改组* < b id = "洗牌" > < / b >  
＊  
*  Reducer从每个  
* {@link Mapper}使用HTTP跨网络  
李* < / >  
＊  
李* < >  
* < b id = "排序" > < / b >排序  
＊  
* 
框架合并排序<代码>Reducer输入  
* <代码>关键代码> < / s  
*(因为不同的Mappers可能输出相同的键)  
＊  
* 
 shuffle和sort阶段同时发生，即当输出是  
 . txt >  . txt  
＊  
* < b id = " SecondarySort " > SecondarySort < / b >  
＊  
* 对返回的值进行二级排序  
*迭代器时，应用程序应该使用secondary扩展键  
键并定义一个分组比较器。 索引键将被排序  
*整个键，但将使用分组比较器分组决定  
在同一个reduce调用中发送哪些键和值。 分组  
*比较器通过  
* {@link工作# setGroupingComparatorClass(类)}。 排序顺序是  
*控制  
* {@link工作# setSortComparatorClass(类)}。< / p >  
 
Reducer核心调用
  
大致处理流程
  
提交任务执行流程
 
 
  
注意：上节 我们讲述了 任务对提交流程 
Mapreduce执行机制之提交任务和切片原理
  
这次看看Map和Reduce执行流程 
我们的Driver 类 
public class WordCountDriver {

    public static void main(String[] args) throws IOException, ClassNotFoundException, InterruptedException {
        // 获取job实例
        Job job = Job.getInstance();

        // 设置 驱动类
        job.setJarByClass(WordCountDriver.class);

        // 关联map和reducer
        job.setMapperClass(WordCountMapper.class);
        job.setReducerClass(WordCountReducer.class);
        // 设置预聚合combiner
        job.setCombinerClass(WordCountCombiner.class);
        System.out.println();

        // 设置map输出
        job.setMapOutputKeyClass(Text.class);
        job.setMapOutputValueClass(LongWritable.class);

        // 设置分区类，自定义按条件将数据输出分区
        job.setPartitionerClass(CustomPartitioner.class);
        // 设置ReduceTask数量
        job.setNumReduceTasks(4);

        // 设置总输出
        job.setOutputKeyClass(Text.class);
        job.setOutputValueClass(LongWritable.class);

        // 设置文件的输出路径和输出路径
        FileInputFormat.setInputPaths(job,"G:\input\wordCountInput");
        FileOutputFormat.setOutputPath(job,new Path("G:\output\wordOutput11"));

        // 提交job
        System.exit(job.waitForCompletion(true) ? 0 : 1);
    }
}
 
partitioner 自定义分区类 

 Combiner 预结合类 
 
Mapper 类 
 
reducer 类 
 
如何执行？ 

 
  
走到自定义的Mapper
 
 
 
 
  

  
重点： 从这里我们可以看到，partitioner的执行时机是在MapTask中context.write()之后 ,将数据输出到collector环形缓存区之前，确定数据的分区，reduceTask之前 
现在进入环形缓冲区collector 
 public synchronized void collect(K key, V value, final int partition
                                     ) throws IOException {
      reporter.progress();

	  **  校验keyClass是否与定义的一致
      if (key.getClass() != keyClass) {
        throw new IOException("Type mismatch in key from map: expected "
                              + keyClass.getName() + ", received "
                              + key.getClass().getName());
      }

	  **  校验ValueClass是否与定义的一致
      if (value.getClass() != valClass) {
        throw new IOException("Type mismatch in value from map: expected "
                              + valClass.getName() + ", received "
                              + value.getClass().getName());
      }

	  **  判断partitioner分区的合法性
      if (partition < 0 || partition >= partitions) {
        throw new IOException("Illegal partition for " + key + " (" +
            partition + ")");
      }

	  ** 校验是否出现spill溢写异常
      checkSpillException();
      bufferRemaining -= metaSIZE;
      
      ** 判断环形缓冲区， 缓存是否80M已经用完，如果用完，开始溢写
      if (bufferRemaining <= 0) {
        // start spill if the thread is not running and the soft limit has been
		**  使用可重入锁ReentrantLock 锁住这段代码
        // reached   
        spillLock.lock();
        try {
          do {
            if (!spillInProgress) {
              final int kvbidx = 4 * kvindex;
              final int kvbend = 4 * kvend;
              // serialized, unspilled bytes always lie between kvindex and
              // bufindex, crossing the equator. Note that any void space
              // created by a reset must be included in "used" bytes
              final int bUsed = distanceTo(kvbidx, bufindex);
              final boolean bufsoftlimit = bUsed >= softLimit;
              if ((kvbend + metaSIZE) % kvbuffer.length !=
                  equator - (equator % metaSIZE)) {
                // spill finished, reclaim space
                resetSpill();
                bufferRemaining = Math.min(
                    distanceTo(bufindex, kvbidx) - 2 * metaSIZE,
                    softLimit - bUsed) - metaSIZE;
                continue;
              } else if (bufsoftlimit && kvindex != kvend) {
                // spill records, if any collected; check latter, as it may
                // be possible for metadata alignment to hit spill pcnt
                startSpill();
                final int avgRec = (int)
                  (mapOutputByteCounter.getCounter() /
                  mapOutputRecordCounter.getCounter());
                // leave at least half the split buffer for serialization data
                // ensure that kvindex >= bufindex
                final int distkvi = distanceTo(bufindex, kvbidx);
                final int newPos = (bufindex +
                  Math.max(2 * metaSIZE - 1,
                          Math.min(distkvi / 2,
                                   distkvi / (metaSIZE + avgRec) * metaSIZE)))
                  % kvbuffer.length;
                setEquator(newPos);
                bufmark = bufindex = newPos;
                final int serBound = 4 * kvend;
                // bytes remaining before the lock must be held and limits
                // checked is the minimum of three arcs: the metadata space, the
                // serialization space, and the soft limit
                bufferRemaining = Math.min(
                    // metadata max
                    distanceTo(bufend, newPos),
                    Math.min(
                      // serialization max
                      distanceTo(newPos, serBound),
                      // soft limit
                      softLimit)) - 2 * metaSIZE;
              }
            }
          } while (false);
        } finally {

			** 释放锁
          spillLock.unlock();
        }
      }

      try {
        
        ** 序列化key
        // serialize key bytes into buffer
        int keystart = bufindex;
        keySerializer.serialize(key);
        if (bufindex < keystart) {
          // wrapped the key; must make contiguous
          bb.shiftBufferedKey();
          keystart = 0;
        }
        ** 序列化value
        // serialize value bytes into buffer
        final int valstart = bufindex;
        valSerializer.serialize(value);
        // It's possible for records to have zero length, i.e. the serializer
        // will perform no writes. To ensure that the boundary conditions are
        // checked and that the kvindex invariant is maintained, perform a
        // zero-length write into the buffer. The logic monitoring this could be
        // moved into collect, but this is cleaner and inexpensive. For now, it
        // is acceptable.
        
        ** 将数据写入缓冲区
        bb.write(b0, 0, 0);

        // the record must be marked after the preceding write, as the metadata
        // for this record are not yet written
        int valend = bb.markRecord();

        mapOutputRecordCounter.increment(1);
        mapOutputByteCounter.increment(
            distanceTo(keystart, valend, bufvoid));

        // write accounting info
        kvmeta.put(kvindex + PARTITION, partition);
        kvmeta.put(kvindex + KEYSTART, keystart);
        kvmeta.put(kvindex + VALSTART, valstart);
        kvmeta.put(kvindex + VALLEN, distanceTo(valstart, valend));
        // advance kvindex
        kvindex = (kvindex - Nmeta + kvmeta.capacity()) % kvmeta.capacity();
      } catch (MapBufferTooSmallException e) {
        LOG.info("Record too large for in-memory buffer: " + e.getMessage());
        spillSingleRecord(key, value, partition);
        mapOutputRecordCounter.increment(1);
        return;
      }
    }
 
1、bufferRemaining 默认80M，环形缓冲区，溢写阈值
  
mapTask循环将数据写入到环形缓冲区之后，自定义map走完，关闭环形缓冲区 

 
 
 
  
private void sortAndSpill() throws IOException, ClassNotFoundException,
                                       InterruptedException {
      //approximate the length of the output file to be the length of the
      //buffer + header lengths for the partitions
      final long size = distanceTo(bufstart, bufend, bufvoid) +
                  partitions * APPROX_HEADER_LENGTH;
      FSDataOutputStream out = null;
      FSDataOutputStream partitionOut = null;
      try {
        // create spill file
        final SpillRecord spillRec = new SpillRecord(partitions);
        final Path filename =
            mapOutputFile.getSpillFileForWrite(numSpills, size);
        out = rfs.create(filename);

        final int mstart = kvend / Nmeta;
        final int mend = 1 + // kvend is a valid record
          (kvstart >= kvend
          ? kvstart
          : kvmeta.capacity() + kvstart) / Nmeta;
        sorter.sort(MapOutputBuffer.this, mstart, mend, reporter);
        int spindex = mstart;
        final IndexRecord rec = new IndexRecord();
        final InMemValBytes value = new InMemValBytes();
        for (int i = 0; i < partitions; ++i) {
          IFile.Writer writer = null;
          try {
            long segmentStart = out.getPos();
            partitionOut = CryptoUtils.wrapIfNecessary(job, out, false);
            writer = new Writer(job, partitionOut, keyClass, valClass, codec,
                                      spilledRecordsCounter);

			** 判断是否定义了Combiner，我们自己定义了
            if (combinerRunner == null) {
              // spill directly
              DataInputBuffer key = new DataInputBuffer();
              while (spindex < mend &&
                  kvmeta.get(offsetFor(spindex % maxRec) + PARTITION) == i) {
                final int kvoff = offsetFor(spindex % maxRec);
                int keystart = kvmeta.get(kvoff + KEYSTART);
                int valstart = kvmeta.get(kvoff + VALSTART);
                key.reset(kvbuffer, keystart, valstart - keystart);
                getVBytesForOffset(kvoff, value);
                writer.append(key, value);
                ++spindex;
              }
            } else {
              int spstart = spindex;
              while (spindex < mend &&
                  kvmeta.get(offsetFor(spindex % maxRec)
                            + PARTITION) == i) {
                ++spindex;
              }
              // Note: we would like to avoid the combiner if we've fewer
              // than some threshold of records for a partition
              if (spstart != spindex) {
                combineCollector.setWriter(writer);
                RawKeyValueIterator kvIter =
                  new MRResultIterator(spstart, spindex);
				
				**  走自定义的combiner
                combinerRunner.combine(kvIter, combineCollector);
              }
            }

            // close the writer
            writer.close();
            if (partitionOut != out) {
              partitionOut.close();
              partitionOut = null;
            }

            // record offsets
            rec.startOffset = segmentStart;
            rec.rawLength = writer.getRawLength() + CryptoUtils.cryptoPadding(job);
            rec.partLength = writer.getCompressedLength() + CryptoUtils.cryptoPadding(job);
            spillRec.putIndex(rec, i);

            writer = null;
          } finally {
            if (null != writer) writer.close();
          }
        }

        if (totalIndexCacheMemory >= indexCacheMemoryLimit) {
          // create spill index file
          Path indexFilename =
              mapOutputFile.getSpillIndexFileForWrite(numSpills, partitions
                  * MAP_OUTPUT_INDEX_RECORD_LENGTH);
          spillRec.writeToFile(indexFilename, job);
        } else {
          indexCacheList.add(spillRec);
          totalIndexCacheMemory +=
            spillRec.size() * MAP_OUTPUT_INDEX_RECORD_LENGTH;
        }
        LOG.info("Finished spill " + numSpills);
        ++numSpills;
      } finally {
        if (out != null) out.close();
        if (partitionOut != null) {
          partitionOut.close();
        }
      }
    }
 
走到我们自定义的Combiner 

  
 
重点：我们从这里可以看到combiner的执行时机： 环形缓冲区的数据 排序溢写 到 本地临时文件之前，注意：此时数据已经按Key排序好了，数据溢写到本地文件之前，对数据进行一个提前预聚合combiner。 
 
注意：如果使用了combiner，就不要使用自定义的Reducer不然会导致最后没有数据 
 
接下来走到我们自定义Reducer 

 initialize
 
 
  
第一阶段 COPY
 
  
第二阶段sort
  
第三阶段调用reduce
 
 JobRunner就是我们之前确定的clientProtocol，它分别调用mapTask的run()，和ReduceTask的Run()
 
 
 
 
  
最终输出的结果
 
 我们定义了4个分区，一个分区一个ReduceTask处理数据，而一个reduceTask会输出到一个文件，因此4个文件 

 
 
  
在看看我们的输入文件，统计单词个数
  
整个流程如下：

Mapreduce执行机制之Map和Reduce

大数据系统相关栏目本月热门文章