HBase操作命令总结

1，如何运行HBase

1，如何查找hbase的安装目录

whereis用来查找程序的安装目录、帮助文档等等，如下：

whereis hbase

结果如下，目录下一级包含bin的就是hbase的安装目录。
[hadoop@cdh002 bin]$ whereis hbase
hbase: /usr/bin/hbase /etc/hbase /usr/lib/hbase

2，如何运行hbase命令

首先切换到hbase的安装目录
cd /usr/lib/hbase/bin/

执行hbase shell
hbase shell

3 如何查找当前用户

使用whoami命令可查看当前用户

hbase(main):001:0> whoami
hadoop (auth:SIMPLE)
groups: hadoop

4 HBase如何删除字符

需要同时按住Ctrl + Backspace 才能删除之前输入的字符。

2，如何操作HBase表结构

1，查看有哪些表

list

2,如何创建HBase表

创建HBase表不需要指定列，只需要指定列簇就行了
语法：create <table>, {NAME => <family>, VERSIONS => <VERSIONS>}
比如创建一个表，保存用户的个人信息infos和工作经历jobs
(1)
hbase(main):007:0> create "user",{ NAME => "infos",VERSIONS => 1},{NAME => "jobs"}
0 row(s) in 4.6440 seconds

=> Hbase::Table - user

注意：NAME要大写，否则报错提示未定义的本地变量
NameError: undefined local variable or method `name' for #<Object:0x79972a95>

(2)简单的写法：表名,列簇
create "user","infos","jobs"

3，如何查看HBase的表结构

desc "user" 或者 describe "user"

hbase(main):008:0> describe "user"
Table user is ENABLED
user
COLUMN FAMILIES DESCRIPTION
{NAME => 'infos', DATA_BLOCK_ENCODING => 'NONE', BLOOMFILTER => 'ROW', REPLICATION_SCOPE => '0'
, VERSIONS => '1', COMPRESSION => 'NONE', MIN_VERSIONS => '0', TTL => 'FOREVER', KEEP_DELETED_C
ELLS => 'FALSE', BLOCKSIZE => '65536', IN_MEMORY => 'false', BLOCKCACHE => 'true'}
{NAME => 'jobs', DATA_BLOCK_ENCODING => 'NONE', BLOOMFILTER => 'ROW', REPLICATION_SCOPE => '0',
VERSIONS => '1', COMPRESSION => 'NONE', MIN_VERSIONS => '0', TTL => 'FOREVER', KEEP_DELETED_CE
LLS => 'FALSE', BLOCKSIZE => '65536', IN_MEMORY => 'false', BLOCKCACHE => 'true'}
2 row(s) in 0.0400 seconds

表结构信息显示有两个列簇

4，如何修改HBase的表结构

(1)首先要禁用表
disable "user"
(2)
删除字段指定METHOD => "delete"
因为有3台机器，因此更新了3台机器 Updating all regions with the new schema

hbase(main):002:0> alter "user",{NAME => "infos"},{NAME => "jobs",METHOD => "delete"},{NAME => "educations"}
Updating all regions with the new schema...
1/1 regions updated.
Done.
Updating all regions with the new schema...
1/1 regions updated.
Done.
Updating all regions with the new schema...
1/1 regions updated.
Done.
0 row(s) in 8.0190 seconds

(3)然后启用表
enable 'user'

5，如何删除HBase表

要先禁用表
hbase(main):004:0> disable "user"
0 row(s) in 0.0410 seconds

再drop表
hbase(main):005:0> drop "user"
0 row(s) in 1.3550 seconds

3，如何操作HBase表数据

1，hbase如何插入数据

语法：put <table>,<rowkey>,<family:column>,<value>
put 表名,列簇:列,值

hbase(main):020:0> put "user","rowno001","infos:name","zhangsan"
0 row(s) in 0.1400 seconds

hbase(main):021:0> put "user","rowno001","infos:age","18"
0 row(s) in 0.0100 seconds

注意：
(1)一次只能插入一列的值；同一行的多列，要分开列插入

2，查询记录

(1)查看所有记录
scan "user"
数据量很大时，千万不要用这个命令
(2) 查找前2条数据
hbase(main):028:0> scan "user",{LIMIT => 2}
ROW COLUMN+CELL
rowno001 column=infos:age, timestamp=1528170157838, value=18
rowno001 column=infos:name, timestamp=1528170127697, value=zhangsan
rowno002 column=infos:age, timestamp=1528170336721, value=19

(3)查找某范围内的记录
hbase(main):032:0> scan "user",{ STARTROW => "rowno002", ENDROW => "rowno004"}
ROW COLUMN+CELL
rowno002 column=infos:age, timestamp=1528170336721, value=19
rowno003 column=infos:name, timestamp=1528171140729, value=xiaosan

注意：ENDROW是不包含的。

(4)查找某一行记录
hbase(main):034:0> get "user","rowno001"
COLUMN CELL
infos:age timestamp=1528170157838, value=18
infos:name timestamp=1528170127697, value=zhangsan
2 row(s) in 0.0250 seconds

可以指定列名
hbase(main):035:0> get "user","rowno001","infos:name"
COLUMN CELL
infos:name timestamp=1528170127697, value=zhangsan

3，统计记录

(1)统计总共多少行
hbase(main):033:0> count "user"
3 row(s) in 0.0290 seconds
=> 3

4，修改一条数据

和添加数据的命令是一致的，修改一下值就行，会直接覆盖
put "user","rowno003","infos:name","san"

5，删除数据

删除某一列：
delete "user","rowno003","infos:name"

删除某一行：
注意删除某一行，用的是deleteall
deleteall "user","rowno002"