安装

上传解压HBase安装包

tar -xvzf hbase-2.1.0.tar.gz -C ../server/

修改HBase配置文件

1. hbase-env.sh

cd /export/server/hbase-2.1.0/conf
vim hbase-env.sh
# 第28行
export JAVA_HOME=/export/server/jdk1.8.0_241/
export HBASE_MANAGES_ZK=false

2. hbase-site.xml
   vim hbase-site.xml

hbase.rootdirhdfs://node1.itcast.cn:8020/hbasehbase.cluster.distributedtruehbase.zookeeper.quorumnode1.itcast.cn,node2.itcast.cn,node3.itcast.cnhbase.zookeeper.property.dataDir/export/server/apache-zookeeper-3.6.0-bin/datahbase.unsafe.stream.capability.enforcefalse

配置环境变量

1. 配置Hbase环境变量

vim /etc/profile
export HBASE_HOME=/export/server/hbase-2.1.0
export PATH=$PATH:${HBASE_HOME}/bin:${HBASE_HOME}/sbin

2. 加载环境变量
   source /etc/profile

复制jar包到lib

cp $HBASE_HOME/lib/client-facing-thirdparty/htrace-core-3.1.0-incubating.jar $HBASE_HOME/lib/

修改regionservers文件

vim regionservers
node1.itcast.cn
node2.itcast.cn
node3.itcast.cn

分发安装包与配置文件

cd /export/server
scp -r hbase-2.1.0/ node2.itcast.cn:$PWD
scp -r hbase-2.1.0/ node3.itcast.cn:$PWD
scp -r /etc/profile node2.itcast.cn:/etc
scp -r /etc/profile node3.itcast.cn:/etc

在node2.itcast.cn和node3.itcast.cn加载环境变量
source /etc/profile

启动HBase

cd /export/onekey

1. 启动ZK
   ./start-zk.sh
2. 启动hadoop
   start-dfs.sh
3. 启动hbase
   start-hbase.sh

验证Hbase是否启动成功

1. 启动hbase shell客户端
   hbase shell
2. 输入status

[root@node1 onekey]# hbase shell
SLF4J: Class path contains multiple SLF4J bindings.
SLF4J: Found binding in [jar:file:/export/server/hadoop-2.7.5/share/hadoop/common/lib/slf4j-log4j12-1.7.10.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: Found binding in [jar:file:/export/server/hbase-2.1.0/lib/client-facing-thirdparty/slf4j-log4j12-1.7.25.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation.
SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory]
HBase Shell
Use "help" to get list of supported commands.
Use "exit" to quit this interactive shell.
Version 2.1.0, re1673bb0bbfea21d6e5dba73e013b09b8b49b89b, Tue Jul 10 17:26:48 CST 2018
Took 0.0034 seconds                                                                                                                                           
Ignoring executable-hooks-1.6.0 because its extensions are not built. Try: gem pristine executable-hooks --version 1.6.0
Ignoring gem-wrappers-1.4.0 because its extensions are not built. Try: gem pristine gem-wrappers --version 1.4.0
2.4.1 :001 > status
1 active master, 0 backup masters, 3 servers, 0 dead, 0.6667 average load
Took 0.4562 seconds                                                                                                                                           
2.4.1 :002 >

Web UI

安装目录说明

参考硬件配置

针对大概800TB存储空间的集群中每个Java进程的典型内存配置：

推荐：

1. Master机器要运行NameNode、ResourceManager、以及HBase HMaster，推荐24GB左右
2. Slave机器需要运行DataNode、NodeManager和HBase RegionServer，推荐24GB（及以上）
3. 根据CPU的核数来选择在某个节点上运行的进程数，例如：两个4核CPU=8核，每个Java进程都可以独立占有一个核（推荐：8核CPU）
4. 内存不是越多越好，在使用过程中会产生较多碎片，Java堆内存越大， 会导致整理内存需要耗费的时间越大。例如：给RegionServer的堆内存设置为64GB就不是很好的选择，一旦FullGC就会造成较长时间的等待，而等待较长，Master可能就认为该节点已经挂了，然后移除掉该节点