HBase 1.2.4 部署

1. 上传hbase-1.2.4-bin.tar.gz并解压

# mkdir /root/Hbase

# cd /root/Hbase

# rz

通过弹出的“打开”对话框从宿主机上选择已经下载好的hbase-1.2.4-bin.tar.gz

# tar -xzvf hbase-1.2.4-bin.tar.gz

2. 配置profile

# vim /etc/profile

在文件最后添加如下配置

export HBASE_HOME=/root/Hbase/hbase-1.2.4

export PATH=$PATH:$HBASE_HOME/bin

# source /etc/profile

3. Lib更新

# cd /root/Hbase/hbase-1.2.4/lib

执行如下命令

# cp ~/Hbase/hadoop-2.7.3/share/hadoop/mapreduce/lib/hadoop-annotations-2.7.3.jar.

# cp~/Hbase/hadoop-2.7.3/share/hadoop/tools/lib/hadoop-auth-2.7.3.jar .

# cp~/Hbase/hadoop-2.7.3/share/hadoop/common/hadoop-common-2.7.3.jar .

# cp~/Hbase/hadoop-2.7.3/share/hadoop/hdfs/hadoop-hdfs-2.7.3.jar .

 

# cp~/Hbase/hadoop-2.7.3/share/hadoop/mapreduce/hadoop-mapreduce-client-app-2.7.3.jar.

# cp~/Hbase/hadoop-2.7.3/share/hadoop/mapreduce/hadoop-mapreduce-client-common-2.7.3.jar.

# cp~/Hbase/hadoop-2.7.3/share/hadoop/mapreduce/hadoop-mapreduce-client-core-2.7.3.jar.

# cp~/Hbase/hadoop-2.7.3/share/hadoop/mapreduce/hadoop-mapreduce-client-jobclient-2.7.3.jar.

# cp~/Hbase/hadoop-2.7.3/share/hadoop/mapreduce/hadoop-mapreduce-client-shuffle-2.7.3.jar.

 

# cp ~/Hbase/hadoop-2.7.3/share/hadoop/yarn/hadoop-yarn-api-2.7.3.jar.

# cp~/Hbase/hadoop-2.7.3/share/hadoop/yarn/hadoop-yarn-client-2.7.3.jar .

# cp~/Hbase/hadoop-2.7.3/share/hadoop/yarn/hadoop-yarn-common-2.7.3.jar .

# cp~/Hbase/hadoop-2.7.3/share/hadoop/yarn/hadoop-yarn-server-common-2.7.3.jar .

 

# 解决java.lang.NoClassDefFoundError: org/htrace/Trace

# cp~/Hbase/hadoop-2.7.3/share/hadoop/common/lib/htrace-core-3.0.4.jar .

 

# 删除老版的jar

rm *-2.5.1.jar

4. 配置hbase-env.sh

# vim /root/Hbase /hbase-1.2.4/conf/hbase-env.sh

修改如下配置

export JAVA_HOME=/usr/java/jdk1.8.0_112/

export HBASE_MANAGES_ZK=true             #使用自带的zookeeper

exportHBASE_CLASSPATH=/root/Hadoop/hadoop-2.7.3/etc/hadoop

# 注释掉下面的配置（因为1.8JDK没有这个选项）

# exportHBASE_MASTER_OPTS="$HBASE_MASTER_OPTS -XX:PermSize=128m-XX:MaxPermSize=128m"

#export HBASE_REGIONSERVER_OPTS="$HBASE_REGIONSERVER_OPTS-XX:PermSize=128m -XX:MaxPermSize=128m"

5. 配置hbase-site.xml

# vim /root/Hbase/hbase-1.2.4/conf/hbase-site.xml

在<configuration></configuration>中加入以下内容

       <property>

              <name>hbase.rootdir</name>

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

              <description>hadoop集群地址</description> 

       </property>

       <property>

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

              <value>true</value>

              <description>是否启动集群模式</description>

       </property>

       <property>

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

              <value>/root/Hbase/hbase-1.2.4/tmp</value>

       </property>

       <property>

              <name>hbase.master</name>

              <value>master:60000</value>

              <description> 指定hbase集群主控节点</description>

       </property>

       <property>

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

              <value>slave1,slave2,slave3</value>

              <description>zookeeper集群主机名列表</description>

       </property>

    <property>

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

        <value>/root/Hbase/hbase-1.2.4/zookeeper_data</value>

    </property>

6. 配置regionservers

# vim $HBASE_HOME/conf/regionservers

修改为如下设置

master

slave1

slave2

slave3

注意：master必须添加，否则执行start-hbase.sh会报错

7. 拷贝hbase到其他节点

把hadoop的hdfs-site.xml和core-site.xml 放到hbase/conf下

cp/root/Hadoop/hadoop-2.7.3/etc/hadoop/hdfs-site.xml/root/Hbase/hbase-1.2.4/conf/

cp /root/Hadoop/hadoop-2.7.3/etc/hadoop/core-site.xml/root/Hbase/hbase-1.2.4/conf/

scp -r ~/Hbase/hbase-1.2.4 slave1:~/Hbase

scp -r ~/Hbase/hbase-1.2.4 slave2:~/Hbase

scp -r ~/Hbase/hbase-1.2.4 slave3:~/Hbase

8. 启动验证

启动

# start-hbase.sh

通过浏览器访问hbaseHMaster Web页面

http://slave1:16010

HRegionServer Web页面

http://slave1:16030

http://slave2:16030

http://slave3:16030

shell验证

# hbase shell

list验证

hbase(main):001:0> list

建表验证

hbase(main):001:0> create 'user','name','sex'

9. 创建bigdata表

hbase(main):017:0> create 'bigdata',{NAME=> 'labels', VERSIONS => 2},{NAME => 'fields', VERSIONS => 2}

10. 遇到的困难及解决方法

10.1 执行start-hbase.sh启动hbase出现警告Java HotSpot(TM) 64-Bit ServerVM warning: ignoring option PermSize=128m;如下图

由于JDK使用的是jdk1.8.0_65

$ vim $HBASE_HOME/conf/hbase-env.sh

注释掉如下两行

#exportHBASE_MASTER_OPTS="$HBASE_MASTER_OPTS -XX:PermSize=128m-XX:MaxPermSize=128m"

#exportHBASE_REGIONSERVER_OPTS="$HBASE_REGIONSERVER_OPTS -XX:PermSize=128m-XX:MaxPermSize=128m"

10.2 Hbase启动后不久，HMaster和HRetionServer自动关闭

原因一：防火墙没有关闭，解决办法参考本章第2节。

原因二：系统时间不一致，解决办法：修改系统时间，比如：# date -s 00:00:00

或者依次输入如下命令，按提示操作：

$ tzselect

$ TZ='Asia/Shanghai'; export TZ

$ clock --set --date="08/31/17 14:37"

$ clock --hctosys

10.3 执行hbase shell命令时提示如下警告

SLF4J: Class path contains multiple SLF4Jbindings.

SLF4J: Found binding in[jar:file:/root/Hbase/hbase-1.2.4/lib/slf4j-log4j12-1.7.5.jar!/org/slf4j/impl/StaticLoggerBinder.class]

SLF4J: Found binding in[jar:file:/root/Hadoop/hadoop-2.7.3/share/hadoop/common/lib/slf4j-log4j12-1.7.10.jar!/org/slf4j/impl/StaticLoggerBinder.class]

slf4j-log4j重复包含，解决办法：将habase-1.2.4/lib/slf4j-log4j12-1.7.5.jar删除。注意，如果删除的是$HADOOP_HOME/ share/hadoop/common/lib/slf4j-log4j12-1.7.10.jar，那么启动和停止HDFS时会报如下警告：

10.4 启动HBase后执行hbase(main):001:0> list ，报错：org.apache.hadoop.hbase.PleaseHoldException:Master is initializing如图：

quit之后，重新hbase shell就好了

10.5 启动HBase后执行hbase(main):001:0> list ，报错：***Server is not running yet

如下图：

解决办法：

# stop-hbase.sh

# hadoop dfsadmin -safemode leave

# hbase shell