Mac下hadoop运行word count的坑

Mac下hadoop运行word count的坑

Word count体现了Map Reduce的经典思想，是分布式计算中中的hello world。然而博主很幸运地遇到了Mac下特有的问题Mkdirs failed to create，特此记录

一、代码

1. WCMapper.java

package wordcount;

import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.util.StringUtils;

import java.io.IOException;

/**
 * 四个泛型中，前两个是指mapper输入的数据类型
 * KEYIN是输入的key类型，VALUEIN是输入的value类型
 * map和reduce的数据输入输出都是以key-value对的形式分装的
 * 默认情况下，框架传递给我们的mapper的输入数据中
 * key是要处理的文本中第一行的起始偏移量，value是这一行的内容
 *
 * Long->LongWritable实现hadoop自己的序列化接口，内容更精简，传输效率高
 * String->Text
 */
public class WCMapper extends Mapper<LongWritable, Text, Text, LongWritable>{
    //mapreduce框架每一行数据就调用一次改方法
    @Override
    protected void map(LongWritable key, Text value, Context context) throws IOException, InterruptedException {
        // 具体的业务逻辑就写在这个方法中，而且需要的处理的key-value已经传递进来
        // 将这一行的内容转换成string
        String line = value.toString();
        // 切分单词
        String[] words = StringUtils.split(line, ' ');
        // 通过context把结果输出
        for (String word: words){
            context.write(new Text(word), new LongWritable(1));
        }
    }
}

2. WCReducer.java

package wordcount;

import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Reducer;

import java.io.IOException;

public class WCReducer extends Reducer<Text, LongWritable, Text, LongWritable>{
    // 框架在map处理完成之后，将所有k-v对缓存起来
    // 进行分组，然后传递一个组<key, values{}>
    // 调用一次reduce方法
    @Override
    protected void reduce(Text key, Iterable<LongWritable> values, Context context) throws IOException, InterruptedException {
        long count = 0;
        // 遍历values，累加求和
        for (LongWritable value: values){
            count += value.get();
        }
        // 输出这一个单词的统计结果
        context.write(key, new LongWritable(count));
    }
}

3. WCRunner.java(启动项)

package wordcount;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

import java.io.IOException;


/**
 * 用来描述一个特定的作业
 * 比如，该作业使用哪个类作为逻辑处理的map，哪个作为reduce
 * 还可以指定该作业要需要的数据所在的路径
 * 还可以指定该作业输出的结果放到哪个路径
 */
public class WCRunner {
    public static void main(String[] args) throws IOException, ClassNotFoundException, InterruptedException {
        Configuration conf = new Configuration();
        Job job = Job.getInstance(conf);

        // 设置整个job需要的jar包
        // 通过WCRuner来找到其他依赖WCMapper和WCReducer
        job.setJarByClass(WCRunner.class);

        // 本job使用的mapper和reducer类
        job.setMapperClass(WCMapper.class);
        job.setReducerClass(WCReducer.class);

        // 指定reducer的输出kv类型
        job.setOutputKeyClass(Text.class);
        job.setOutputValueClass(LongWritable.class);

        // 指定mapper的输出kv类型
        job.setMapOutputKeyClass(Text.class);
        job.setMapOutputValueClass(LongWritable.class);

        // 指定原始数据存放在哪里
        FileInputFormat.setInputPaths(job,new Path("/wc/input/"));
        // 指定处理结果的输出数据存放在哪里
        FileOutputFormat.setOutputPath(job, new Path("/wc/output/"));

        // 将job提交运行
        job.waitForCompletion(true);

    }
}

二、问题重现

写好代码后打包成jar，博主是用IDEA直接图形化操作的，然后提交到hadoop上运行

hadoop jar hadoopStudy.jar wordcount.WCRunner

结果未像官网和其他很多教程中说的那样出结果，而是报错

Exception in thread "main" java.io.IOException: Mkdirs failed to create /var/folders/vf/rplr8k812fj018q5lxcb5k940000gn/T/hadoop-unjar1598612687383099338/META-INF/license
	at org.apache.hadoop.util.RunJar.ensureDirectory(RunJar.java:146)
	at org.apache.hadoop.util.RunJar.unJar(RunJar.java:119)
	at org.apache.hadoop.util.RunJar.unJar(RunJar.java:94)
	at org.apache.hadoop.util.RunJar.run(RunJar.java:227)
	at org.apache.hadoop.util.RunJar.main(RunJar.java:153)

最后折腾了半天，发现是Mac的问题，在stackoverflow中找到解释

The issue is that a /tmp/hadoop-xxx/xxx/LICENSE file and a
/tmp/hadoop-xxx/xxx/license directory are being created on a
case-insensitive file system when unjarring the mahout jobs.

删除原来压缩包的META-INF/LICENS，再重新压缩，解决问题~

zip -d hadoopStudy.jar META-INF/LICENSE
jar tvf hadoopStudy.jar | grep LICENSE

然后把新的jar上传到hadoop上运行

hadoop jar hadoopStudy.jar wordcount.WCRunner

bingo!

三、运行结果

顺便用浏览器看一下运行结果

• 输入文件wc/input/input.txt

• 输出文件/wc/output/part-r-00000]

运行结果显然是正确的，再也不敢随便说Mac大法好了……