hadoop-HBase-observer的一个例子，hadoophbase

文章由LinuxBoy分享于2019-03-27 06:03:36热评（559）

hadoop-HBase-observer的一个例子，hadoophbase

hbase(main):021:0> describe 'users'
DESCRIPTION ENABLED
{NAME => 'users', coprocessor$1 => 'file:///home/u/myjar/UsersObserver.jar|test.hbase.inaction.example5_2.Users true
Observer|1001|', FAMILIES => [{NAME => 'info', BLOOMFILTER => 'NONE', REPLICATION_SCOPE => '0', COMPRESSION => 'N
ONE', VERSIONS => '3', TTL => '2147483647', MIN_VERSIONS => '0', BLOCKSIZE => '65536', IN_MEMORY => 'false', BLOCK
CACHE => 'true'}]}
1 row(s) in 0.0610 seconds

先disable表，然后alert一下，那个1001是优先级，别忘记enable

disable 'users'
alter 'users', METHOD => 'table_att', 'coprocessor'=>'file:///home/u/mylib/UsersObserver.jar|test.hbase.inaction.example5_2.UsersObserver|1001|'
alter 'users', METHOD => 'table_att_unset',NAME => 'coprocessor$2' 删除
enable 'users'

hbase(main):022:0> scan 'users'
ROW COLUMN+CELL
id01 column=info:email, timestamp=1413963413002, value=wyj@gmail.com
id01 column=info:name, timestamp=1413963413002, value=wyj
id01 column=info:password, timestamp=1413963413002, value=000000
id01 column=info:user, timestamp=1413963413002, value=id01
id09 column=info:email, timestamp=1414566775616, value=test9@gmail.com
id09 column=info:name, timestamp=1414566775616, value=test9
id09 column=info:password, timestamp=1414566775616, value=9
id09 column=info:user, timestamp=1414566775616, value=id09
id99 column=info:email, timestamp=1414565339530, value=test99@gmail.com
id99 column=info:name, timestamp=1414565339530, value=test99
id99 column=info:password, timestamp=1414565339530, value======01=====
id99 column=info:user, timestamp=1414565339530, value=id99
invalID column=info:invalid_pass, timestamp=1414566775657, value=invalid_pass:9
11 row(s) in 0.1290 seconds

start() 和stop()这两个方法，在表enable和disable时分别也会对应执行

这start()和stop()在连接池里获得连接与释放连接实在是多余

observer类似触发器，操作一张表的时候要注意递归调用，拦截到一次put后再put这个表就会递归

这个例子就是这样的，实际上没任何意义，就是个例子而已

注意锁，HBase是行级的锁，如果想put同一行是不会成功的

observer很不好调试，只能整一堆的log......

package test.hbase.inaction.example5_2;

import java.io.IOException;

import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.hbase.CoprocessorEnvironment;
import org.apache.hadoop.hbase.client.HTableInterface;
import org.apache.hadoop.hbase.client.HTablePool;
import org.apache.hadoop.hbase.client.Put;
import org.apache.hadoop.hbase.coprocessor.BaseRegionObserver;
import org.apache.hadoop.hbase.coprocessor.ObserverContext;
import org.apache.hadoop.hbase.coprocessor.RegionCoprocessorEnvironment;
import org.apache.hadoop.hbase.regionserver.wal.WALEdit;
import org.apache.hadoop.hbase.util.Bytes;

import test.hbase.inaction.example2_4.User;
import test.hbase.inaction.example2_4.UsersDAO;

/*
  <span style="color:#ff0000;">先disable表，然后alert一下，那个1001是优先级，别忘记enable</span>
 disable 'users'
 alter 'users', METHOD => 'table_att', 'coprocessor'=>'file:///home/u/mylib/FollowsObserver.jar|test.hbase.inaction.example5_2.FollowsObserver|1001|'
 alter 'users', METHOD => 'table_att_unset',NAME => 'coprocessor$2'<span style="white-space:pre">	</span>
 enable 'users'
 describe 'users'
 */
public class UsersObserver extends BaseRegionObserver {

	private final Log log = LogFactory.getLog(this.getClass());

	private HTablePool pool = null;

	public void start(CoprocessorEnvironment env) throws IOException {
		log.info("WYJ ---------------------------- start() begin");
		pool = new HTablePool(env.getConfiguration(), Integer.MAX_VALUE);
		log.info("pool=" + pool);
		log.info("WYJ ---------------------------- start() end");
	}

	public void stop(CoprocessorEnvironment env) throws IOException {
		log.info("WYJ ---------------------------- stop() begin");
		pool.close();
		log.info("WYJ ---------------------------- stop() end");
	}

	public void postPut(final ObserverContext<RegionCoprocessorEnvironment> e,
			final Put put, final WALEdit edit, final boolean writeToWAL) {
		log.info("WYJ: ================================================ postPut() begin");
		try {
			byte[] table = e.getEnvironment().getRegion().getRegionInfo()
					.getTableName();
			log.info("table=" + Bytes.toString(table));
			if (!Bytes.equals(table, UsersDAO.TABLE_NAME)) {
				return;
			}
			String pass = Bytes.toString(put.get(UsersDAO.INFO_FAM, UsersDAO.PASS_COL).get(0).getValue());
			log.info("pass=" + pass);
			byte[] user = put.get(UsersDAO.INFO_FAM, UsersDAO.USER_COL).get(0).getValue();
			log.info("user=" + Bytes.toString(user));
			if (pass == null || pass.length() < 6) {
				log.info("password.length=" + pass.length());
				UsersDAO dao = new UsersDAO(pool);
				log.info("pool=" + pool);
				dao.addUser("id99", "test99", "test99@gmail.com", "=====01=====");
				log.info("dao.addUser(\"id99\", \"test99\", \"test99@gmail.com\", \"99999999\"); FINAL");
			}
		} catch (Exception ex) {
			log.error(ex, ex);
		}
		log.info("WYJ: ================================================ postPut() end");
	}
}

在hadoop环境下运行一个wordcount例子

首先进入hadoop安装目录下，新建一个hdfs下的目录，./bin/hadoop fs -mkdir /input，如果你有了，就不必了。接着上传你要进行wordcount的文件到hdfs中，比如这个文件是test.txt，命令是./bin/hadoop fs -put test.txt /input/。然后执行./bin/hadoop jar hadoop-0.20.2-examples.jar wordcount /input /output，然后等执行完，在hdsf下的/output路径下找到结果就好了。
你说的权限问题，你执行 ls -l . 查看是不是你文件的所有者或者群组不一致了，可以用命令chown 或者chgrp来改动。

举例出一个类似“顶针”的例子，并冠以诸如“一生的戒指”之类的名称用30--50字解释所蕴含的感情

时光刻画了年轮，年轮蹉跎了岁月，岁月抚慰了无数不安的日子；在那些不安的日子里，感谢时时在梦中出现的你；而你，就是整个世界。

推荐文章：

hadoop-HBase-observer的一个例子，hadoophbase