DistributedCache的使用方法（hadoop0.20.2）

public class ProcedureJob {        public static class MapClass extends  Mapper<Object, Text, Text, Text>{                private Path[] localFiles;                private URI[] uris;                @Override                protected void setup(Context context) throws IOException,                                InterruptedException {                       localFiles = DistributedCache.getLocalCacheFiles(context.getConfiguration());                       uris = DistributedCache.getCacheFiles(context.getConfiguration());                       System.out.println(uris[0].toString()+"+++++++++++++++"+localFiles+"----------------");                       if(uris==null || uris.length==0){                               throw new IOException("Error reading file from distributed cache. No URIs found.");                       }                       FileSystem fs = FileSystem.get(URI.create("hdfs://namenode:9000"),context.getConfiguration());                       Path localPath = new Path("~/abc.txt");                       fs.copyToLocalFile(new Path(uris[0].toString()),localPath );                       BufferedReader br = new BufferedReader(new FileReader(localPath.toString()));                       String str="";                       while((str=br.readLine())!=null){                               System.out.println(str+"---------------------------------------------");                       }                       br.close();                }                protected void map(Object key, Text value, Context context)                                throws IOException, InterruptedException {                      context.write(new Text(key.toString()), new Text("1"));                }        }                public static void main(String[] args) throws Exception {                Configuration conf = new Configuration();                                DistributedCache.addCacheFile(new Path(args[0]).toUri(), conf);                                  Job job = new Job(conf,"Test");                job.setJarByClass(ProcedureJob.class);                FileInputFormat.setInputPaths(job,new Path(args[1]));                FileOutputFormat.setOutputPath(job, new Path(args[2]));                job.setJobName("Test");                job.setMapperClass(MapClass.class);                job.setOutputKeyClass(Text.class);                job.setOutputValueClass(Text.class);                job.setNumReduceTasks(0);                job.waitForCompletion(true);        }}

说明：事先将本地文件上传到HDFS中，/data/pubic/test.txt。

注意点：
1.Job对象必须在使用configuration对象把缓存文件添加之后才能new，否则会无法添加，出现java.lang.NullPointerException。
2.在setup中获取Path[] localFiles = DistributedCache.getLocalCacheFiles(context.getConfiguration());得到的对象是null，可能版本问题。
3.在setup方法中获取文件系统,FileSystem fs = FileSystem.get(URI.create("hdfs://namenode:9000"),context.getConfiguration());
如果直接是FileSystem fs = FileSystem.get(context.getConfiguration());会出现java.lang.IllegalArgumentException: Wrong FS: hdfs://localhost:54310/user/Hadoop/b, expected: file:/// 异常。