【大数据系列】windows环境下搭建hadoop开发环境使用api进行基本操作

前言

搭建完hadoop集群之后在windows环境下搭建java项目进行测试 操作hdfs中的文件

版本一

package com.slp.hadoop274.hdfs;

import java.io.IOException;
import java.io.InputStream;
import java.net.URL;
import java.net.URLConnection;

import org.apache.hadoop.fs.FsUrlStreamHandlerFactory;
import org.junit.Test;

/**
 *
 * @author sangliping
 *完成hdfs操作
 */
public class TestHDFS {

    /**
     * 读取hdfs文件
     * @throws IOException
     */
	@Test
	public void readFile() throws IOException{
		URL url = new URL("hdfs://192.168.181.201:8020/user/sanglp/hadoop/copyFromLocal");
		URLConnection con = url.openConnection();
		InputStream is = con.getInputStream();
		byte[] buf = new byte[is.available()];
		is.read(buf);
		is.close();
		String str = new String(buf,"UTF-8");
		System.out.println(str);
	}
}

　　以上运行测试的时候会报错，原因是URL无法识别hdfs协议。

版本二、

package com.slp.hadoop274.hdfs;

import java.io.IOException;
import java.io.InputStream;
import java.net.URL;
import java.net.URLConnection;

import org.apache.hadoop.fs.FsUrlStreamHandlerFactory;
import org.junit.Test;

/**
 *
 * @author sangliping
 *完成hdfs操作
 */
public class TestHDFS {

	static{
		//注册hdfs协议否则URL无法识别该协议
		URL.setURLStreamHandlerFactory(new FsUrlStreamHandlerFactory());
	}
    /**
     * 读取hdfs文件
     * @throws IOException
     */
	@Test
	public void readFile() throws IOException{
		URL url = new URL("hdfs://192.168.181.201:8020/user/sanglp/hadoop/copyFromLocal");
		URLConnection con = url.openConnection();
		InputStream is = con.getInputStream();
		byte[] buf = new byte[is.available()];
		is.read(buf);
		is.close();
		String str = new String(buf,"UTF-8");
		System.out.println(str);
	}
}

　　这个时候就可以正确的打印出hdfs文件copyFromLocal的文件内容。

附：可以将hadoop解压文件下etc中的log4j.properties文件放到项目文件src文件下使控制台打印更友好。

版本三

        /**
	 * 通过hadoop  api读取文件
	 * @throws IOException
	 * java.lang.IllegalArgumentException: Wrong FS: hdfs://192.168.181.201:8020/user/sanglp/hadoop/copyFromLocal, expected: file:///
	 * at org.apache.hadoop.fs.FileSystem.checkPath(FileSystem.java:649)
	 */
	@Test
	public void readFileByApiWrong() throws IOException{
		Configuration con = new Configuration();
		FileSystem fs = FileSystem.get(con);
		Path p = new Path("hdfs://192.168.181.201:8020/user/sanglp/hadoop/copyFromLocal");
		FSDataInputStream fis = fs.open(p);
		ByteArrayOutputStream baos = new ByteArrayOutputStream();
		byte [] buf = new byte[1024];
		int len = -1;
		while((len=fis.read(buf))!=-1){
			baos.write(buf,0,len);
		}
		fis.close();
		baos.close();
		System.out.println(new String(baos.toByteArray(),"UTF-8"));
	}

　　此版本错误，因为未指定namenode

版本四

        /**
	 * 使用API用传统流读取hadoop文件
	 * @throws IOException
	 */
	@Test
	public void readFileByApi() throws IOException{
		Configuration con = new Configuration();
		con.set("fs.defaultFS", "hdfs://192.168.181.201:8020");
		FileSystem fs = FileSystem.get(con);
		//以下两种设置path的方法都可以
		//Path p = new Path("hdfs://192.168.181.201:8020/user/sanglp/hadoop/copyFromLocal");
		Path p = new Path("/user/sanglp/hadoop/copyFromLocal");

		FSDataInputStream fis = fs.open(p);
		ByteArrayOutputStream baos = new ByteArrayOutputStream();
		byte [] buf = new byte[1024];
		int len = -1;
		while((len=fis.read(buf))!=-1){
			baos.write(buf,0,len);
		}
		fis.close();
		baos.close();

		System.out.println(new String(baos.toByteArray(),"UTF-8"));
	}

　　版本五

        /**
	 * 使用API并用hadoop提供的IO工具读取hadoop文件
	 * @throws IOException
	 */
	@Test
	public void readFileByApiUsUtils() throws IOException{
		Configuration con = new Configuration();
		con.set("fs.defaultFS", "hdfs://192.168.181.201:8020");
		FileSystem fs = FileSystem.get(con);
		//以下两种设置path的方法都可以
		//Path p = new Path("hdfs://192.168.181.201:8020/user/sanglp/hadoop/copyFromLocal");
		Path p = new Path("/user/sanglp/hadoop/copyFromLocal");

		FSDataInputStream fis = fs.open(p);
		ByteArrayOutputStream baos = new ByteArrayOutputStream();
		int buf = 1024;
		IOUtils.copyBytes(fis, baos, buf);
		System.out.println(new String(baos.toByteArray(),"UTF-8"));
	}

　　版本六

/**
	 * 使用API创建文件夹
	 * @throws IOException
	 * org.apache.hadoop.security.AccessControlException: Permission denied: user=hadoop, access=WRITE, inode="/user/sanglp":sanglp:supergroup:drwxr-xr-x
     * Caused by: org.apache.hadoop.ipc.RemoteException(org.apache.hadoop.security.AccessControlException): Permission denied: user=hadoop, access=WRITE, inode="/user/sanglp":sanglp:supergroup:drwxr-xr-x
	 */
	@Test
	public void makeDir() throws IOException{
		Configuration con = new Configuration();
		con.set("fs.defaultFS", "hdfs://192.168.181.201:8020");
		FileSystem fs = FileSystem.get(con);
		fs.mkdirs(new Path("/user/sanglp/myhadoop"));
	}

　　直接使用上诉API会出现没有权限的问题，需要修改权限

hadoop fs -chmod 777 /user/sanglp

　　版本七

 /**
     * 使用API创建文件
     * @throws IOException
     */
    @Test
    public void putFile() throws IOException{
        Configuration con = new Configuration();
        con.set("fs.defaultFS", "hdfs://192.168.181.201:8020");
        FileSystem fs = FileSystem.get(con);
       FSDataOutputStream out =  fs.create(new Path("/user/sanglp/myhadoop/a.txt"));
       out.write("test put file on myhadoop ".getBytes());
       out.close();
    }