进入hbase shell console

[hadoop@mylab2 hbase]$ ./bin/hbase shell
2016-09-18 15:01:36,773 WARN  [main] util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
SLF4J: Class path contains multiple SLF4J bindings.
SLF4J: Found binding in [jar:file:/usr/local/hbase/lib/slf4j-log4j12-1.7.5.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: Found binding in [jar:file:/usr/local/hadoop/share/hadoop/common/lib/slf4j-log4j12-1.7.10.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; enter 'help<RETURN>' for list of supported commands.
Type "exit<RETURN>" to leave the HBase Shell
Version 1.2.1, r8d8a7107dc4ccbf36a92f64675dc60392f85c015, Wed Mar 30 11:19:21 CDT 2016

hbase(main):001:0>

查询服务器状态

hbase(main):007:0> status
1 active master, 0 backup masters, 2 servers, 0 dead, 1.5000 average load

查看hbase版本

hbase(main):008:0> version
1.2.1, r8d8a7107dc4ccbf36a92f64675dc60392f85c015, Wed Mar 30 11:19:21 CDT 2016


命令帮助

hbase(main):002:0> help get

ERROR: wrong number of arguments (0 for 2)

Here is some help for this command:
Get row or cell contents; pass table name, row, and optionally
a dictionary of column(s), timestamp, timerange and versions. Examples:

  hbase> get 'ns1:t1', 'r1'
  hbase> get 't1', 'r1'
  hbase> get 't1', 'r1', {TIMERANGE => [ts1, ts2]}
  hbase> get 't1', 'r1', {COLUMN => 'c1'}
  hbase> get 't1', 'r1', {COLUMN => ['c1', 'c2', 'c3']}
  hbase> get 't1', 'r1', {COLUMN => 'c1', TIMESTAMP => ts1}
  hbase> get 't1', 'r1', {COLUMN => 'c1', TIMERANGE => [ts1, ts2], VERSIONS => 4}
  hbase> get 't1', 'r1', {COLUMN => 'c1', TIMESTAMP => ts1, VERSIONS => 4}
  hbase> get 't1', 'r1', {FILTER => "ValueFilter(=, 'binary:abc')"}
  hbase> get 't1', 'r1', 'c1'
  hbase> get 't1', 'r1', 'c1', 'c2'
  hbase> get 't1', 'r1', ['c1', 'c2']
  hbase> get 't1', 'r1', {COLUMN => 'c1', ATTRIBUTES => {'mykey'=>'myvalue'}}
  hbase> get 't1', 'r1', {COLUMN => 'c1', AUTHORIZATIONS => ['PRIVATE','SECRET']}
  hbase> get 't1', 'r1', {CONSISTENCY => 'TIMELINE'}
  hbase> get 't1', 'r1', {CONSISTENCY => 'TIMELINE', REGION_REPLICA_ID => 1}

Besides the default 'toStringBinary' format, 'get' also supports custom formatting by
column.  A user can define a FORMATTER by adding it to the column name in the get
specification.  The FORMATTER can be stipulated: 

 1. either as a org.apache.hadoop.hbase.util.Bytes method name (e.g, toInt, toString)
 2. or as a custom class followed by method name: e.g. 'c(MyFormatterClass).format'.

Example formatting cf:qualifier1 and cf:qualifier2 both as Integers: 
  hbase> get 't1', 'r1' {COLUMN => ['cf:qualifier1:toInt',
    'cf:qualifier2:c(org.apache.hadoop.hbase.util.Bytes).toInt'] } 

Note that you can specify a FORMATTER by column only (cf:qualifier).  You cannot specify
a FORMATTER for all columns of a column family.
    
The same commands also can be run on a reference to a table (obtained via get_table or
create_table). Suppose you had a reference t to table 't1', the corresponding commands
would be:

  hbase> t.get 'r1'
  hbase> t.get 'r1', {TIMERANGE => [ts1, ts2]}
  hbase> t.get 'r1', {COLUMN => 'c1'}
  hbase> t.get 'r1', {COLUMN => ['c1', 'c2', 'c3']}
  hbase> t.get 'r1', {COLUMN => 'c1', TIMESTAMP => ts1}
  hbase> t.get 'r1', {COLUMN => 'c1', TIMERANGE => [ts1, ts2], VERSIONS => 4}
  hbase> t.get 'r1', {COLUMN => 'c1', TIMESTAMP => ts1, VERSIONS => 4}
  hbase> t.get 'r1', {FILTER => "ValueFilter(=, 'binary:abc')"}
  hbase> t.get 'r1', 'c1'
  hbase> t.get 'r1', 'c1', 'c2'
  hbase> t.get 'r1', ['c1', 'c2']
  hbase> t.get 'r1', {CONSISTENCY => 'TIMELINE'}
  hbase> t.get 'r1', {CONSISTENCY => 'TIMELINE', REGION_REPLICA_ID => 1}


HBase Shell, version 1.2.1, r8d8a7107dc4ccbf36a92f64675dc60392f85c015, Wed Mar 30 11:19:21 CDT 2016
Type 'help "COMMAND"', (e.g. 'help "get"' -- the quotes are necessary) for help on a specific command.
Commands are grouped. Type 'help "COMMAND_GROUP"', (e.g. 'help "general"') for help on a command group.

COMMAND GROUPS:
  Group name: general
  Commands: status, table_help, version, whoami

  Group name: ddl
  Commands: alter, alter_async, alter_status, create, describe, disable, disable_all, drop, drop_all, enable, enable_all, exists, get_table, is_disabled, is_enabled, list, locate_region, show_filters

  Group name: namespace
  Commands: alter_namespace, create_namespace, describe_namespace, drop_namespace, list_namespace, list_namespace_tables

  Group name: dml
  Commands: append, count, delete, deleteall, get, get_counter, get_splits, incr, put, scan, truncate, truncate_preserve

  Group name: tools
  Commands: assign, balance_switch, balancer, balancer_enabled, catalogjanitor_enabled, catalogjanitor_run, catalogjanitor_switch, close_region, compact, compact_rs, flush, major_compact, merge_region, move, normalize, normalizer_enabled, normalizer_switch, split, trace, unassign, wal_roll, zk_dump

  Group name: replication
  Commands: add_peer, append_peer_tableCFs, disable_peer, disable_table_replication, enable_peer, enable_table_replication, list_peers, list_replicated_tables, remove_peer, remove_peer_tableCFs, set_peer_tableCFs, show_peer_tableCFs

  Group name: snapshots
  Commands: clone_snapshot, delete_all_snapshot, delete_snapshot, list_snapshots, restore_snapshot, snapshot

  Group name: configuration
  Commands: update_all_config, update_config

  Group name: quotas
  Commands: list_quotas, set_quota

  Group name: security
  Commands: grant, list_security_capabilities, revoke, user_permission

  Group name: procedures
  Commands: abort_procedure, list_procedures

  Group name: visibility labels
  Commands: add_labels, clear_auths, get_auths, list_labels, set_auths, set_visibility

SHELL USAGE:
Quote all names in HBase Shell such as table and column names.  Commas delimit
command parameters.  Type <RETURN> after entering a command to run it.
Dictionaries of configuration used in the creation and alteration of tables are
Ruby Hashes. They look like this:

  {'key1' => 'value1', 'key2' => 'value2', ...}

and are opened and closed with curley-braces.  Key/values are delimited by the
'=>' character combination.  Usually keys are predefined constants such as
NAME, VERSIONS, COMPRESSION, etc.  Constants do not need to be quoted.  Type
'Object.constants' to see a (messy) list of all constants in the environment.

If you are using binary keys or values and need to enter them in the shell, use
double-quote'd hexadecimal representation. For example:

  hbase> get 't1', "key\x03\x3f\xcd"
  hbase> get 't1', "key\003\023\011"
  hbase> put 't1', "test\xef\xff", 'f1:', "\x01\x33\x40"

The HBase shell is the (J)Ruby IRB with the above HBase-specific commands added.
For more on the HBase Shell, see http://hbase.apache.org/book.html


表的管理

创建表

hbase(main):004:0> create'scores','grade', 'course'
0 row(s) in 2.8930 seconds

=> Hbase::Table - scores
#查看创建表
hbase(main):006:0> list
TABLE                                                                                                                                
scores                                                                                                                               
1 row(s) in 0.0270 seconds

=> ["scores"]

查看表的构造

hbase(main):010:0> describe'scores'
Table scores is ENABLED                                                                                                              
scores                                                                                                                               
COLUMN FAMILIES DESCRIPTION                                                                                                          
{NAME => 'course', DATA_BLOCK_ENCODING => 'NONE', BLOOMFILTER => 'ROW', REPLICATION_SCOPE => '0', VERSIONS => '1', COMPRESSION => 'NO
NE', MIN_VERSIONS => '0', TTL => 'FOREVER', KEEP_DELETED_CELLS => 'FALSE', BLOCKSIZE => '65536', IN_MEMORY => 'false', BLOCKCACHE => 
'true'}                                                                                                                              
{NAME => 'grade', DATA_BLOCK_ENCODING => 'NONE', BLOOMFILTER => 'ROW', REPLICATION_SCOPE => '0', VERSIONS => '1', COMPRESSION => 'NON
E', MIN_VERSIONS => '0', TTL => 'FOREVER', KEEP_DELETED_CELLS => 'FALSE', BLOCKSIZE => '65536', IN_MEMORY => 'false', BLOCKCACHE => '
true'}                                                                                                                               
2 row(s) in 0.2090 seconds

hbase(main):011:0>

加入一行数据,行名称为zkb 列族grad的列名为”” 值位5

hbase(main):011:0> put'scores','zkb','grade:','5' 
0 row(s) in 0.1240 seconds

hbase(main):012:0>

给zkb这一行的数据的列族course添加一列<math,97>

hbase(main):012:0> put'scores','zkb','course:math','97'
0 row(s) in 0.0120 seconds

hbase(main):013:0>

给zkb这一行的数据的列族course添加一列<art,87>

hbase(main):013:0> put'scores','zkb','course:art','87'
0 row(s) in 0.0140 seconds

hbase(main):014:0>


加入一行数据,行名称为baoniu 列族grad的列名为”” 值为4

hbase(main):014:0>  put'scores','baoniu','grade:','4'
0 row(s) in 0.0170 seconds


给baoniu这一行的数据的列族course添加一列<math,89>　　

hbase(main):015:0> put'scores','baoniu','course:math','89'
0 row(s) in 0.0120 seconds

给Jerry这一行的数据的列族course添加一列<art,80>

hbase(main):016:0> put'scores','baoniu','course:art','80'
0 row(s) in 0.0150 seconds

hbase(main):017:0>


查看scores表中zkb的相关数据

hbase(main):017:0> get'scores','zkb'
COLUMN                             CELL                                                                                              
 course:art                        timestamp=1474183749925, value=87                                                                 
 course:math                       timestamp=1474183718550, value=97                                                                 
 grade:                            timestamp=1474183615369, value=5                                                                  
3 row(s) in 0.1280 seconds

hbase(main):018:0>

查看scores表中所有数据

hbase(main):018:0> scan'scores'
ROW                                COLUMN+CELL                                                                                       
 baoniu                            column=course:art, timestamp=1474183803493, value=80                                              
 baoniu                            column=course:math, timestamp=1474183793034, value=89                                             
 baoniu                            column=grade:, timestamp=1474183782970, value=4                                                   
 zkb                               column=course:art, timestamp=1474183749925, value=87                                              
 zkb                               column=course:math, timestamp=1474183718550, value=97                                             
 zkb                               column=grade:, timestamp=1474183615369, value=5                                                   
2 row(s) in 0.0590 seconds

hbase(main):019:0>

 查看scores表中所有数据courses列族的所有数据

hbase(main):019:0> scan'scores',{COLUMNS => 'course'}
ROW                                COLUMN+CELL                                                                                       
 baoniu                            column=course:art, timestamp=1474183803493, value=80                                              
 baoniu                            column=course:math, timestamp=1474183793034, value=89                                             
 zkb                               column=course:art, timestamp=1474183749925, value=87                                              
 zkb                               column=course:math, timestamp=1474183718550, value=97                                             
2 row(s) in 0.0330 seconds

hbase(main):020:0>

参考文档

http://blog.csdn.net/lifuxiangcaohui/article/details/40042117

http://blog.csdn.net/smcwwh/article/details/7468672

http://www.cnblogs.com/nexiyi/p/hbase_shell.html