7.hbase shell命令 cmd
$HADOOP_USER_NAME
#创建命名空间
create_namespace 'bd1902'
#展示所有命名空间
list_namespace
#删除命名空间,The namespace must be empty.
drop_namespace 'IMUT'
create 't1', 'f1', 'f2', 'f3'
create 't1', {NAME => 'f1'}, {NAME => 'f2'}, {NAME => 'f3'}
#创建一张表，指定版本号为3
create 'bd1902:student', 'baseinfo', 'extrainfo'
create 'bd1902:student1', {NAME => 'baseinfo', VERSIONS => 3},{NAME => 'extrainfo',VERSIONS => 5}
create 'bd1803:employee', 'baseinfo', 'extrainfo'
create 'bd1803:employee1', {NAME => 'baseinfo', VERSIONS => 3},{NAME => 'extrainfo',VERSIONS => 5}
describe 'bd1902:student2'
describe 'bigdata:test1'
hbase 热点问题，数据倾斜
读操作 写操作
1. 默认分区
2. rowkey递增
解决热点问题：
1 预分区 （建表过程中）
2 随机产生rowkey hash 、MD5 、 SHA256
#创建表，预定义分区，在rowkey为0<= <10 10<= 20 20<= 30
create 'bd1902:student2', {NAME=>'baseinfo',VERSIONS=>3}, SPLITS => ['1000', '2000', '3000', '4000'] 四个分界点，分五个区
create 'bd1803:employee3', {NAME=>'baseinfo',VERSIONS=>3}, SPLITS => ['1000', '2000', '3000', '4000']
put 'hbase_test:teacher3','2000009','baseinfo:name','zhangsan'
#创建表，分区标准在文件中，如果rowkey以0001等开头,进行分区使用| 或者 ~ 帮助划分rowkey区域，文件放在进入hbase shell 的目录下
create 'bd1902:student3','baseinfo',{SPLITS_FILE => '/home/briup/splits.txt'}
测试ROWKEY开闭区间：(左闭右开)
put 'bd1902:student2','1000','baseinfo:name' ,'jack'
create 'bd1803:employee4','baseinfo',{SPLITS_FILE => '/home/hbase/sps.txt'}
create 'bd1803:employee4', 'baseinfo', SPLITS_FILE => 'sps.txt'
#使用HexStringSplit算法进行分区，分成10个region，适合散列字符不包含中文，适合16进制的rowkey或者前缀是16进制的rowkey （哈希算法、SHA32）
create 'bd1902:student4', 'baseinfo', {NUMREGIONS => 10, SPLITALGO => 'HexStringSplit'}
create 'bd1803:employee5', 'baseinfo', {NUMREGIONS => 10, SPLITALGO => 'HexStringSplit'}
#使用UniformSplit算法进行分区，rowkey可以包含中文，适合随机字节数组rowkey
create 'bd1902:student5', 'baseinfo', {NUMREGIONS => 5, SPLITALGO => 'UniformSplit'}
create 'bd1803:employee6', 'baseinfo', {NUMREGIONS => 5, SPLITALGO => 'UniformSplit'}
put 'bd1803:employee3', '2000', 'baseinfo:name', '张三'
JRuby （脚本）变量
#create 返回引用值
t1 = create 't1', 'f1'
#alter修改表结构--增加列族
alter 'bd1902:student' ,{NAME => 'extrainfo' ,VERSIONS => 5},{NAME => 'secret',VERSIONS => 5 }
alter 'bigdata:test2', {NAME => 'extrainfo', IN_MEMORY => true}, {NAME => 'secret', VERSIONS => 5}
#alter修改表结构--删除列族
alter 'bd1902:student',{METHOD => 'delete',NAME => 'baseinfo'}
alter 'bigdata:test2', {METHOD => 'delete',NAME => 'baseinfo'}
---------------------------------------
#插入数据 兼顾更新
Cell
put 'ns:t','r','cf:q','v'[,'t']
put 'bd1902:student1','1001','baseinfo:name','Kenvin'
put 'bd1902:student1','1001','baseinfo:gender','male'
put 'bd1902:student1','1001','baseinfo:age','40'
put 'bd1902:student1','1001','baseinfo:pos','CTO'
put 'bd1902:student1','1002','baseinfo:name','Terry'
put 'bd1902:student1','1002','baseinfo:gender','male'
put 'bd1902:student1','1002','baseinfo:age','36'
put 'bd1902:student1','1002','baseinfo:pos','Manager'
put 'bd1902:student1','2001','baseinfo:name','Wood'
put 'bd1902:student1','2001','baseinfo:gender','male'
put 'bd1902:student1','2001','baseinfo:age','32'
put 'bd1902:student1','2001','baseinfo:pos','Manager'
put 'bd1902:student1','2002','baseinfo:name','Terry'
put 'bd1902:student1','2002','baseinfo:gender','male'
put 'bd1902:student1','2002','baseinfo:age','30'
put 'bd1902:student1','2002','baseinfo:pos','Teacher'
put 'bd1902:student1','3001','baseinfo:name','Lurry'
put 'bd1902:student1','3001','baseinfo:gender','male'
put 'bd1902:student1','3001','baseinfo:age','36'
put 'bd1902:student1','3001','baseinfo:pos','Teacher'
scan 'bd1902:student1'
put 'bd1803:employee1','1001','baseinfo:gender','male'
put 'briup:employee3','2000','baseinfo:name','tom'
#插入指定timestamp
put 'hbase_test:teacher5','100000000','extrainfo:salary','5000',1488888888888
#查询
get 单行查询
scan 多行查询
#获得某一个特定值
get 't1', 'r1', ['c1', 'c2']
get 'bigdata:test1','10','baseinfo:name'
#获得前5个版本的数据
get 'bd1803:employee1','1001',{COLUMN => 'baseinfo:position',VERSIONS => 5}
#获得某个时间段数据,不一定是时间最新的数据
get 'hbase_test:teacher2', '10001', {TIMERANGE => [1479371084728, 1479373228331]}
#scan 扫描某张表 select *
scan 'bd1803:employee1'
scan 'bd1902:student1'
#scan 扫描 表中某一列
scan 'test1:student5',{COLUMNS=>'baseinfo:name'}
#scan 使用limit 进行行数限制
scan 'test1:student5',{COLUMNS=>'baseinfo:name',LIMIT=>2}
#scan 指定从某一行开始扫描
scan 'hbase_test:teacher2',{COLUMNS=>'baseinfo:name',LIMIT=>2,STARTROW=>'20001'}
#scan 扫描所有版本
scan 'bigdata:test1','10',{VERSIONS=>5}
#在hbase 对于hfile没有进行过合并操作之前
#scan 超出版本限制也能访问到
scan 'briup:employee3',{VERSIONS=>5,RAW=>true}
#scan 使用过滤器 行键前缀过滤器,只有这一个有属性
scan 'bigdata:test1', {ROWPREFIXFILTER => '10'}
scan 'bd1902:student1', {ROWPREFIXFILTER => '1002'}
#scan 使用空值行健过滤器，只返回行健
scan 'bigdata:test1',{FILTER=>'KeyOnlyFilter()'}
scan 'bigdata:test1',{FILTER=>"ColumnPrefixFilter('na') "}
1 数值 数字
2 CompareFilter.CompareOp 比较符 >
3 ByteArrayComparable binary:1000 substring:
4 byte[] ''
scan 'bd1803:employee1',{FILTER=>"RowFilter(>,'binary:1001')"}
scan 'bd1902:student1',{FILTER=>"RowFilter(>,'binary:2000')"}
#scan 使用行健过滤器,binary: 帮助数据类型转化
scan 'hbase_test:teacher2',{FILTER =>"RowFilter(!=,'binary:10001')"}
#scan 使用列名过滤器
scan 'test1:student5',{FILTER =>"QualifierFilter(>=,'binary:baseinfo:name')"}
#scan 使用子串过滤器
scan 'test1:student5',{FILTER =>"ValueFilter(=,'binary:zhao')"}
#列名前缀过滤器
scan 'test1:student5',{FILTER =>"ColumnPrefixFilter('name')"}
#scan 使用多种过滤器进行条件结合
scan 'hbase_test:teacher2',{FILTER =>"(ValueFilter(=,'binary:hello')) OR (RowFilter (>,'binary:10'))"}
#scan 使用page过滤器，限制每页展示数量
scan 'bigdata:test1',{FILTER =>org.apache.hadoop.hbase.filter.KeyOnlyFilter.new()}
#scan 使用行健过滤器，进行正则表达式的匹配
scan 'test1', {FILTER => RowFilter.new(CompareFilter::CompareOp.valueOf('EQUAL'),RegexStringComparator.new('.*ll.*'))}
scan 'bd1902:student1', {FILTER => org.apache.hadoop.hbase.filter.RowFilter.new(org.apache.hadoop.hbase.filter.CompareFilter::CompareOp.valueOf('EQUAL'),org.apache.hadoop.hbase.filter.RegexStringComparator.new('.*3.*'))}
//-----------------------
#删除数据
delete 't1','r1','c1'
#清空某张表
truncate 't1'
#disable 某张表
disable 'bigdata:test1'
#删除某张表
drop 'bigdata:test2'
#大合并 hfile
major_compact '583b13b5efb36a6ae7794d7e60b4c3a8'
major_compact 'bigdata:test2'
#小合并
#移动region
move 'ENCODED_REGIONNAME', 'SERVER_NAME'
#第一个参数指的是region最后一部分编号(逗号分隔每部分)
move 'a39dc69bd00d19e556ae17e4aeb1ebe1','datanode02,16020,1479354142616'
// 行过滤器
// 1 行健范围
ByteArrayComparable com1 = new BinaryComparator(Bytes.toBytes("briup004"));
RowFilter rf1 = new RowFilter(CompareOp.LESS, com1);
// 2 行健子串范围
ByteArrayComparable com2 = new SubstringComparator("007");
RowFilter rf2 = new RowFilter(CompareOp.EQUAL, com2);
// 3 某个列标示符的值范围
SingleColumnValueFilter scf1 = new SingleColumnValueFilter
(Bytes.toBytes("infos"), Bytes.toBytes("name"), CompareOp.LESS_OR_EQUAL, Bytes.toBytes("张三"));
// 4 匹配正则表达式
ByteArrayComparable com3 = new SubstringComparator("test.");
SingleColumnValueFilter scf2 = new SingleColumnValueFilter
(Bytes.toBytes("infos"), Bytes.toBytes("name"), CompareOp.EQUAL,com3);
// 5 匹配子串 不区分大小写
ByteArrayComparable com4 = new SubstringComparator("te");
SingleColumnValueFilter scf3 = new SingleColumnValueFilter
(Bytes.toBytes("infos"), Bytes.toBytes("name"), CompareOp.EQUAL,com4);
