hbase0.98 fully distributed deployment

	NameNode	DataNode	JournalNode	Zookeeper	Hmaster	HregionServer
node1	1				1
node2	1	1	1	1		1
node3		1	1	1		1
node4		1	1	1	1	1

1, the first extract to the installation package hbase node node1

2, modify the configuration file

hbase-env.sh

export JAVA_HOME=/usr/java/jdk1.8.0_11
export HBASE_MANAGES_ZK=false  #（关闭Hbase本身的zookeeper集群）

hbase-site.xml

<property>
    <name>hbase.rootdir</name>
    <value>hdfs://node2:8020/hbase</value>  <--在HDFS上创建一个干净的节点，用于存放元数据-->
</property>
<property>
    <name>hbase.cluster.distributed</name>
    <value>true</value>  <--是否开启分布式-->
</property>
<property>
    <name>hbase.zookeeper.quorum</name>
    <value>node2,node3,node4</value>  <--对应的zookeeper集群，不用加端口-->
</property>

regionservers from node

node2
node3
node4

backup-masters configuration (Note backup-masters need to manually create) the main backup

node4

3, the hdfs among hdfs-site.xml  copy hbase conf directory (because the required data is stored on hdfs)

cp /opt/hadoop-2.6.5/etc/hadoop/hdfs-site.xml /opt/hbase0.98/conf/

4,  will be distributed to hbase on node1 node2, node3, the current path of node4

scp -r hbase0.98/ root@node2:`pwd`
scp -r hbase0.98/ root@node3:`pwd`
scp -r hbase0.98/ root@node4:`pwd`

5, start

Performed node1 (Master node) start-hbase.sh start the master node.
       In node4 (backup-Master node)  to perform start-hbase.sh start the backup node.