HBase完全分布式安装

假设Hadoop已经成功安装。

实验环境如下：Ubuntu 12
hadoop-0.20.203.0
hbase-0.90.5

三台主机：
192.168.66.91 masternode
192.168.66.92 slavenode1
192.168.66.93 slavenode2

Ubuntu用户名grid
hadoop安装目录：/home/grid/hadoop
hbase计划安装目录：/home/grid/hbase

大致过程为：首先在masternode上安装hbase，配置好后再拷贝至slave机器。
（1）下载hbase-0.90.5.tar.gz，解压到/home/grid/hbase
（2）编辑/home/grid/hbase/conf/hbase-env.sh，编辑后如下所示，带颜色的部分为修改部分：
#
#

# Set environment variables here.

# The java implementation to use.  Java 1.6 required. Java安装目录，最低1.6版本
export JAVA_HOME=/usr/lib/jvm/jdk1.7.0_07

# Extra Java CLASSPATH elements.  Optional. 整合hadoop的配置文件
export HBASE_CLASSPATH=/home/grid/hadoop/conf

# The maximum amount of heap to use, in MB. Default is 1000.
# export HBASE_HEAPSIZE=1000

# Extra Java runtime options.
# Below are what we set by default.  May only work with SUN JVM.
# For more on why as well as other possible settings,
# see http://wiki.apache.org/hadoop/PerformanceTuning
export HBASE_OPTS="-ea -XX:+UseConcMarkSweepGC -XX:+CMSIncrementalMode"

# Uncomment below to enable java garbage collection logging in the .out file.
# export HBASE_OPTS="$HBASE_OPTS -verbose:gc -XX:+PrintGCDetails -XX:+PrintGCDateStamps" 

# Uncomment and adjust to enable JMX exporting
# See jmxremote.password and jmxremote.access in $JRE_HOME/lib/management to configure remote password access.
# More details at: http://java.sun.com/javase/6/doc ... nagement/agent.html
#
# export HBASE_JMX_BASE="-Dcom.sun.management.jmxremote.ssl=false -Dcom.sun.management.jmxremote.authenticate=false"
# export HBASE_MASTER_OPTS="$HBASE_JMX_BASE -Dcom.sun.management.jmxremote.port=10101"
# export HBASE_REGIONSERVER_OPTS="$HBASE_JMX_BASE -Dcom.sun.management.jmxremote.port=10102"
# export HBASE_THRIFT_OPTS="$HBASE_JMX_BASE -Dcom.sun.management.jmxremote.port=10103"
# export HBASE_ZOOKEEPER_OPTS="$HBASE_JMX_BASE -Dcom.sun.management.jmxremote.port=10104"

# File naming hosts on which HRegionServers will run.  $HBASE_HOME/conf/regionservers by default.
# export HBASE_REGIONSERVERS=${HBASE_HOME}/conf/regionservers

# Extra ssh options.  Empty by default.
# export HBASE_SSH_OPTS="-o ConnectTimeout=1 -o SendEnv=HBASE_CONF_DIR"

# Where log files are stored.  $HBASE_HOME/logs by default.  hbase日志目录，注意logs目录需创建
export HBASE_LOG_DIR=/home/grid/hbase/logs

# A string representing this instance of hbase. $USER by default.
# export HBASE_IDENT_STRING=$USER

# The scheduling priority for daemon processes.  See 'man nice'.
# export HBASE_NICENESS=10

# The directory where pid files are stored. /tmp by default.
# export HBASE_PID_DIR=/var/hadoop/pids

# Seconds to sleep between slave commands.  Unset by default.  This
# can be useful in large clusters, where, e.g., slave rsyncs can
# otherwise arrive faster than the master can service them.
# export HBASE_SLAVE_SLEEP=0.1

# Tell HBase whether it should manage it's own instance of Zookeeper or not.开启zookeeper
export HBASE_MANAGES_ZK=true




（3）编辑/home/grid/hbase/conf/hbase-site.xml，默认内无属性，均需自行添加：
   
    hbase.rootdir 
    hdfs://masternode:9000/hbase 
   
   
    hbase.cluster.distributed 
    true 
   
   
    hbase.master 
    hdfs://masternode:6000 
   
   
    hbase.zookeeper.quorum 
    slavenode1 
   
   
    hbase.zookeeper.property.dataDir 
    /home/grid/hbase/zookeeper 



（4）编辑/home/grid/hbase/conf/regionservers，将所有的slavenode添加到这个文件，类似与hadoop中slaves文件，该文件原本为空，如下是我添加的内容，两台slavenode机器的域名（hosts文件中已做解析）：
slavenode1
slavenode2


（5）解决版本、jar包冲突：

1)删除hbase/lib/hadoop-core-0.20-append-r1056497.jar(我是拷贝到其它路径，以备万一)；
2)拷贝hadoop/hadoop-core-0.20.203.0.jar(注意是hadoop根目录下)，hadoop/lib/commons-collections-3.2.1.jar，hadoop/lib/commons-configuration-1.6.jar到hbase/lib下。

（6）将master节点上已经配置好的hbase文件夹拷贝到slave节点，执行：

scp -r /home/grid/hbase grid@slavenode1:/home/grid/hbase

scp -r /home/grid/hbase grid@slavenode2:/home/grid/hbase

（7）启动

1）启动hadoop：/home/grid/hadoop/bin/start-all.sh

执行命令后用jps命令验证，显示：

3549 Jps

2438 NameNode

2758 JobTracker

2690 SecondaryNameNode

说明hadoop启动成功。

2）启动hbase：/home/grid/hbase/bin/start-hbase.sh

执行命令后用jps命令验证，显示：

3750 Jps
3652 HMaster
2438 NameNode
2758 JobTracker
2690 SecondaryNameNode

说明hbase启动成功，然后分别到两台slave机器上执行jps，slavenode1上执行jps，显示：


2676 TaskTracker
2504 DataNode
3800 Jps
3501 HQuorumPeer
3669 HRegionServer



注意：其中HQuorumPeer是zookeeper进程，在hbase-site.xml中的hbase.zookeeper.quorum属性中为slavenode1配置了zookeeper；
slavenode2上执行jps，显示：


2663 TaskTracker
2482 DataNode
3607 Jps
3530 HRegionServer

（8）进入hbase

在master机器执行hbase/bin/hbase shell即可进入hbase的shell

grid@masternode:~$ hbase/bin/hbase shell
HBase Shell; enter 'help' for list of supported commands.
Type "exit" to leave the HBase Shell
Version 0.90.5, r1212209, Fri Dec  9 05:40:36 UTC 2011


hbase(main):001:0>