hbase 集群搭建，

hbase 集群搭建

hbase集群依赖于hdfs,和yarn.所以安装搭建hbase集群必须先完成hdfs和yarn的搭建安装

一 ZooKeeper安装配置

1.在每台服务器(虚拟机)的host中添加(ip +hostname  )：

192.168.15.5 jokeros1

192.168.15.6 jokeros2

192.168.15.7 jokeros3

2.下载并解压zookeeper压缩文件：

tar –zxvf zookeeper-3.4.8.tar.gz

3.配置环境变量：

#zookeeper

export ZOOKEEPER=/opt/SoftWare/zookeeper-3.4.11

export PATH=$PATH:$ZOOKEEPER/bin

4.使修改生效：

source /etc/profile

5.到zookeeper的conf目录下面，新增一个zoo.cfg文件

cp zoo_sample.cfg zoo.cfg

 

6.修改：

dataDir=/usr/tools/zookeeper-3.4.8/data

添加：

server.1=master:2888:3888

server.2=slave1:2888:3888

server.3=slave2:2888:3888

 

7.配置完以后将上述内容全部拷贝到另外两台服务的相同位置

使用scp

scp /usr/tools/zookeeper-3.4.8root@jokeros2: /usr/tools/

scp /usr/tools/zookeeper-3.4.8root@jokeros3: /usr/tools/

 

8.三台机器下面的data目录里面各自建一个myid的文件

然后里面填上相应的数字

如jokeros1是server.1，里面的数字是1

Jokeros2是server.2，里面的数字是2

 

9./etc/profile环境变了也可以用scp来完成，或者可以各自修改成一致的

10.三台分别启动zookeeper

zkServer.sh start

每台机器上查看状态：

zkServer.sh status

结果：

ZooKeeper JMX enabled by default

Using config:/usr/tools/zookeeper-3.4.8/bin/../conf/zoo.cfg

Mode: follower

使用jps查看：

jps

结果

QuorumPeerMain

二 HBase安装配置

1.在某一台上解压hbase的压缩文件，如在192.168.15.5

tar –zxvf hbase-1.2.0-bin.tar.gz

 

2.配置添加环境变量：

#hbase

export HBASE_HOME=/opt/SoftWare/hbase-1.3.1

export PATH=$PATH:$HBASE_HOME/bin

 

3.使环境变量生效

source /etc/profile

 

4.进入hbase的conf目录，需要修改三个文件：hbase-env.sh、hbase-site.xml和regionservers

其中hbase-env.sh中，在文档的十多行位置处添加：

# The java implementation to use.  Java 1.7+ required.

# export JAVA_HOME=/usr/java/jdk1.6.0/

exportJAVA_HOME=/usr/tools/jdk1.8.0_73

# Extra Java CLASSPATH elements.  Optional.

# export HBASE_CLASSPATH=

然后在后面添加：

# Seconds to sleep between slavecommands.  Unset by default.  This

# can be useful in large clusters, where,e.g., slave rsyncs can

# otherwise arrive faster than the mastercan service them.

# export HBASE_SLAVE_SLEEP=0.1

# Tell HBase whether it should manage it'sown instance of Zookeeper or not.

exportHBASE_MANAGES_ZK=false

 

hbase-site.xml中

<configuration>

       <property>

               <name>hbase.zookeeper.quorum</name>

               <value>jokeros1,jokeros2,jokeros3</value>

                <description>Thedirectory shared by RegionServers.</description>

       </property>

       <property>

               <name>hbase.zookeeper.property.dataDir</name>

               <value>/usr/tools/hbase-1.2.0/zookeeperdata</value>

                <description>Propertyfrom ZooKeeper config zoo.cfg.

                The directory where thesnapshot is stored.

                </description>

       </property>

       <property>

               <name>hbase.tmp.dir</name>

               <value>/usr/tools/hbase-1.2.0/tmpdata</value>

       </property>

       <property>

               <name>hbase.rootdir</name>

               <value>hdfs://jokeros1:9000/hbase</value>

                <description>Thedirectory shared by RegionServers.</description>

       </property>

       <property>

               <name>hbase.cluster.distributed</name>

                <value>true</value>

                <description>The mode thecluster will be in. Possible values are

                false: standalone andpseudo-distributed setups with managed Zookeeper

                true: fully-distributed withunmanaged Zookeeper Quorum (see hbase-env.sh)

                </description>

       </property>

</configuration>

5.regionservers文件中添加各个从属服务器的ip或者hostname：

 jokeros1

jokeros2

jokeros3

 

6.保存后分别把hbase的整个文件夹拷贝到其他服务器：

 

scp /usr/tools/hbase-1.2.0 root@jokeros2:/usr/tools/

scp /usr/tools/hbase-1.2.0 root@jokeros3:/usr/tools/

 

7.在hadoop的namenode节点上启动hbase服务

start-hbase.sh

 

启动后：jps

HRegionServer

HMaster

子节点

HRegionServer

 

启动顺序

Hadoop-hdfs-------》hadoop-yarn------》zookeeper------》hbase