hbase 环境部署

1、由于刚搭建hbase，对hbase还不是很熟悉，可能由于hbase和hadoop版本都是分别进行发布和管理的，因此在进行hbase搭建和测试的时候，可能经常出现一些问题，因此建议在搭建环境的时候，建议hbase下所使用的lib库，涉及到hadoop的全部使用hadoop目录下的lib库。
2、安装
    1、准备tar包 hbase-0.94.9.tar.gz
    2、准备环境变量
   [master-/hadoop/hbase/conf]ls
   hadoop-metrics.properties  hbase-env.sh  hbase-policy.xml  hbase-site.xml  log4j.properties  regionservers
    3、修改参数，主要为hbase-env.sh,hbase-site.xml和regionservers几个文件
     修改hbase-env.sh
      export JAVA_HOME=/hadoop/jdk
   export HBASE_HOME=/hadoop/hbase
   export PATH=$PATH:$HBASE_HOME/bin
   export HBASE_OPTS="$HBASE_OPTS -XX:+HeapDumpOnOutOfMemoryError -XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode"
   export HADOOP_HOME=/hadoop/hadoop
     修改hbase-site.xml
     [master-/hadoop/hbase-config]cat hbase-site.xml
     
     
     
    
      
         hbase.rootdir
         hdfs://master:54310/hbase
      
      
         hbase.cluster.distributed
         true
      
      
         hbase.master
         master
      
       
        hbase.zookeeper.quorum 
        master,node1,node2 
       
       
        hbase.tmp.dir 
        /hadoop/hbase/tmp 
       
     
     3、修改regionservers文件
     [master-/hadoop/hbase-config]cat regionservers
          node1                                         
          node2                                         
          master            
          4、设置hbase自己的zookeeper
     [master-/hadoop/hbase-config]cat zoo.cfg
        # The number of milliseconds of each tick
        tickTime=2000
        # The number of ticks that the initial
        # synchronization phase can take
        initLimit=10
        # The number of ticks that can pass between
        # sending a request and getting an acknowledgement
        syncLimit=5
        # the directory where the snapshot is stored.
        # do not use /tmp for storage, /tmp here is just
        # example sakes.
        dataDir=/hadoop/hbase/zookeeper/data
        # the port at which the clients will connect
        clientPort=2181
        #
        # Be sure to read the maintenance section of the
        # administrator guide before turning on autopurge.
        #
        # http://zookeeper.apache.org/doc/current/zookeeperAdmin.html#sc_maintenance
        #
        # The number of snapshots to retain in dataDir
        autopurge.snapRetainCount=3
        # Purge task interval in hours
        # Set to "0" to disable auto purge feature
        autopurge.purgeInterval=1
        
        server.1=node1:2888:3888
        server.2=node2:2888:3888
        server.3=master:2888:3888        
          4、配置文件设置好了
            scp 命令复制到每个节点对应的目录下
            scp hbase* node1:/hadoop/
            scp hbase* node2:/hadoop/
          5、启动hbase
           [master-/hadoop/hbase/bin]./start-hbase.sh
           启动之后，检查每个节点的regionserver日志和zookeeper日志，以及主节点下的master等日志，如果都没有问题，检查hbase是否成功
            检查hbase master
            检查hbase regionserver
     6、检查hbase操作
      hbase(main):001:0> create 'dsg','data'
      0 row(s) in 1.4240 seconds
      hbase(main):003:0> scan 'dsg'
      ROW                                COLUMN+CELL                                                                                     
      0 row(s) in 0.0170 seconds