大数据教程FG074-HBase Shell命令操作实战

本文档风哥主要介绍HBase Shell命令操作实战，包括DDL操作、DML操作、管理操作等内容，风哥教程参考HBase官方文档Shell Commands、Data Model Operations等内容，适合大数据开发运维人员在学习和测试中使用，如果要应用于生产环境则需要自行确认。更多视频教程www.fgedu.net.cn

Part01-基础概念与理论知识

1.1 HBase Shell概述

HBase Shell是HBase提供的命令行交互工具，基于JRuby实现，用于管理和操作HBase数据库。学习交流加群风哥微信: itpux-com

HBase Shell功能：

DDL操作：创建、修改、删除表
DML操作：数据的增删改查
管理操作：集群管理、Region管理
安全操作：用户权限管理
快照操作：快照创建和恢复

1.2 命令分类

HBase Shell命令分类详解：

# HBase Shell命令分类

1. DDL命令
– create：创建表
– alter：修改表
– disable：禁用表
– enable：启用表
– drop：删除表
– describe：查看表结构
– list：列出表

2. DML命令
– put：插入数据
– get：获取数据
– scan：扫描数据
– delete：删除数据
– deleteall：删除整行
– count：统计行数
– append：追加数据
– increment：增量操作

3. 管理命令
– status：集群状态
– balance_switch：负载均衡开关
– balancer：触发负载均衡
– move：移动Region
– split：分裂Region
– merge：合并Region
– major_compact：主压缩
– flush：刷写MemStore

4. 安全命令
– grant：授权
– revoke：撤销权限
– user_permission：查看权限

5. 快照命令
– snapshot：创建快照
– list_snapshots：列出快照
– restore_snapshot：恢复快照
– clone_snapshot：克隆快照
– delete_snapshot：删除快照

1.3 语法规则

HBase Shell语法规则详解：

# HBase Shell语法规则

1. 命令格式
命令 ‘参数1’, ‘参数2’, {属性名 => ‘属性值’}

2. 字符串引号
– 单引号：’value’
– 双引号：”value”
– 推荐使用单引号

3. 列族定义
{NAME => ‘cf’, VERSIONS => 3, COMPRESSION => ‘SNAPPY’}

4. 多参数格式
‘table_name’, {NAME => ‘cf1’}, {NAME => ‘cf2’}

5. Ruby语法
# 变量定义
table_name = ‘fgedu_user’
create table_name, ‘cf1’, ‘cf2’

# 条件判断
if exists?(table_name)
disable table_name
drop table_name
end

# 启动HBase Shell
$ hbase shell

HBase Shell
Use “help” to get list of supported commands.
Use “exit” to quit this interactive shell.
Version 2.5.5, rUnknown, Mon Apr 8 13:00:00 CST 2026

hbase(main):001:0>

# 查看帮助
hbase(main):001:0> help

HBase Shell, version 2.5.5, rUnknown, Mon Apr 8 13:00:00 CST 2026
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: processlist, status, table_help, version, whoami

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

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

风哥提示：HBase Shell支持Ruby语法，可以使用变量和条件判断。建议使用脚本文件执行批量操作。

Part02-生产环境规划与建议

2.1 表设计规划

表设计规划建议：

# 表设计原则

1. 列族数量
– 建议：1-3个
– 原因：每个列族一个Store，影响性能

2. 列族名称
– 简短：info, data, cf
– 有意义：basic, detail, behavior

3. 版本数
– 默认：1
– 根据需求设置：VERSIONS => 3

4. 压缩
– 推荐：SNAPPY
– 高压缩比：GZIP
– 快速：LZ4

5. 预分区
– 根据RowKey设计
– 避免热点

# 表设计示例
create ‘fgedu_user’,
{NAME => ‘basic’, VERSIONS => 1, COMPRESSION => ‘SNAPPY’, BLOOMFILTER => ‘ROW’},
{NAME => ‘behavior’, VERSIONS => 3, COMPRESSION => ‘SNAPPY’, TTL => 2592000},
{NAME => ‘tag’, VERSIONS => 1, COMPRESSION => ‘SNAPPY’},
SPLITS => [‘10000000’, ‘20000000’, ‘30000000’, ‘40000000’, ‘50000000’]

2.2 RowKey设计规划

RowKey设计规划建议：

# RowKey设计原则

1. 唯一性
– 必须唯一标识一行数据

2. 长度
– 建议：10-100字节
– 最大：64KB

3. 散列性
– 避免热点
– 使用散列前缀

4. 有序性
– 支持范围查询
– 根据查询模式设计

# RowKey设计方案

1. 用户ID方案
RowKey: user_id
示例: user_00000001

2. 散列前缀方案
RowKey: MD5(user_id).substring(0,4) + user_id
示例: a1b2_user_00000001

3. 时间戳方案
RowKey: user_id + Long.MAX_VALUE – timestamp
示例: user_00000001_9223372036854775807

4. 组合键方案
RowKey: user_id + ‘_’ + item_id
示例: user_00000001_item_001

# RowKey设计示例
# 用户行为表
RowKey: MD5(user_id).substring(0,4) + ‘_’ + user_id + ‘_’ + timestamp

# 示例数据
a1b2_user_00000001_1680940800000
a1b2_user_00000001_1680940801000
c3d4_user_00000002_1680940800000

2.3 列族设计规划

列族设计规划建议：

# 列族设计原则

1. 相关性
– 相关数据放在同一列族

2. 访问模式
– 经常一起访问的数据放同一列族

3. 数据类型
– 相同类型数据放同一列族

4. TTL
– 根据数据生命周期设置

# 列族属性配置

1. VERSIONS
– 版本数，默认1
– VERSIONS => 3

2. COMPRESSION
– 压缩算法
– COMPRESSION => ‘SNAPPY’

3. BLOOMFILTER
– 布隆过滤器
– BLOOMFILTER => ‘ROW’

4. TTL
– 数据存活时间（秒）
– TTL => 2592000 # 30天

5. BLOCKSIZE
– 块大小
– BLOCKSIZE => 65536 # 64KB

6. IN_MEMORY
– 是否常驻内存
– IN_MEMORY => true

# 列族设计示例
{NAME => ‘basic’, VERSIONS => 1, COMPRESSION => ‘SNAPPY’, BLOOMFILTER => ‘ROW’}
{NAME => ‘behavior’, VERSIONS => 3, COMPRESSION => ‘SNAPPY’, TTL => 2592000}
{NAME => ‘detail’, VERSIONS => 1, COMPRESSION => ‘GZIP’, BLOCKSIZE => 131072}

生产环境建议：生产环境建议合理设计表结构和RowKey，避免热点问题。列族数量不超过3个，启用压缩减少存储空间。学习交流加群风哥QQ113257174

Part03-生产环境项目实施方案

3.1 DDL操作实战

3.1.1 创建表

# 启动HBase Shell
$ hbase shell

# 创建简单表
hbase(main):001:0> create ‘fgedu_user’, ‘info’

Created table fgedu_user

# 创建多列族表
hbase(main):002:0> create ‘fgedu_order’, ‘basic’, ‘detail’, ‘status’

Created table fgedu_order

# 创建带属性的表
hbase(main):003:0> create ‘fgedu_product’,
{NAME => ‘basic’, VERSIONS => 3, COMPRESSION => ‘SNAPPY’, BLOOMFILTER => ‘ROW’},
{NAME => ‘price’, VERSIONS => 1, COMPRESSION => ‘SNAPPY’, TTL => 2592000}

Created table fgedu_product

# 创建预分区表
hbase(main):004:0> create ‘fgedu_behavior’, ‘info’,
SPLITS => [’10’, ’20’, ’30’, ’40’, ’50’, ’60’, ’70’, ’80’, ’90’]

Created table fgedu_behavior

# 使用分区文件创建
hbase(main):005:0> create ‘fgedu_log’, ‘data’, SPLITS_FILE => ‘/tmp/splits.txt’

Created table fgedu_log

# 查看表列表
hbase(main):006:0> list

TABLE
fgedu_behavior
fgedu_log
fgedu_order
fgedu_product
fgedu_user
5 row(s)
Took 0.1234 seconds.
=> [“fgedu_behavior”, “fgedu_log”, “fgedu_order”, “fgedu_product”, “fgedu_user”]

# 查看表结构
hbase(main):007:0> describe ‘fgedu_product’

Table fgedu_product is ENABLED
fgedu_product
COLUMN FAMILIES DESCRIPTION
{NAME => ‘basic’, BLOOMFILTER => ‘ROW’, VERSIONS => ‘3’, IN_MEMORY => ‘false’, KEEP_DELETED_CELLS => ‘FALSE’, DATA_BLOCK_ENCODING => ‘NONE’, TTL => ‘FOREVER’, COMPRESSION => ‘SNAPPY’, MIN_VERSIONS => ‘0’, BLOCKCACHE => ‘true’, BLOCKSIZE => ‘65536’, REPLICATION_SCOPE => ‘0’}
{NAME => ‘price’, BLOOMFILTER => ‘ROW’, VERSIONS => ‘1’, IN_MEMORY => ‘false’, KEEP_DELETED_CELLS => ‘FALSE’, DATA_BLOCK_ENCODING => ‘NONE’, TTL => ‘2592000’, COMPRESSION => ‘SNAPPY’, MIN_VERSIONS => ‘0’, BLOCKCACHE => ‘true’, BLOCKSIZE => ‘65536’, REPLICATION_SCOPE => ‘0’}
2 row(s)
Took 0.0123 seconds.

3.1.2 修改表

# 添加列族
hbase(main):008:0> alter ‘fgedu_user’, ‘detail’

Updating all regions with the new schema…
1/1 regions updated.
Done.
Took 1.2345 seconds.

# 修改列族属性
hbase(main):009:0> alter ‘fgedu_user’, {NAME => ‘info’, VERSIONS => 5}

Updating all regions with the new schema…
1/1 regions updated.
Done.
Took 1.2345 seconds.

# 删除列族
hbase(main):010:0> alter ‘fgedu_user’, {NAME => ‘detail’, METHOD => ‘delete’}

Updating all regions with the new schema…
1/1 regions updated.
Done.
Took 1.2345 seconds.

# 修改表属性
hbase(main):011:0> alter ‘fgedu_user’, MAX_FILESIZE => ‘21474836480’

Updating all regions with the new schema…
1/1 regions updated.
Done.
Took 1.2345 seconds.

# 禁用表
hbase(main):012:0> disable ‘fgedu_user’

Took 0.5678 seconds.

# 启用表
hbase(main):013:0> enable ‘fgedu_user’

Took 0.7890 seconds.

# 检查表状态
hbase(main):014:0> is_enabled ‘fgedu_user’

true
Took 0.0123 seconds.

hbase(main):015:0> is_disabled ‘fgedu_user’

false
Took 0.0123 seconds.

# 删除表
hbase(main):016:0> disable ‘fgedu_test’

Took 0.5678 seconds.

hbase(main):017:0> drop ‘fgedu_test’

Took 0.7890 seconds.

# 检查表是否存在
hbase(main):018:0> exists ‘fgedu_user’

Table fgedu_user does exist
Took 0.0123 seconds.

3.2 DML操作实战

3.2.1 数据插入

# 插入单条数据
hbase(main):019:0> put ‘fgedu_user’, ‘user_00000001’, ‘info:name’, ‘fgedu01’

Took 0.1234 seconds.

hbase(main):020:0> put ‘fgedu_user’, ‘user_00000001’, ‘info:age’, ’25’

Took 0.0123 seconds.

hbase(main):021:0> put ‘fgedu_user’, ‘user_00000001’, ‘info:city’, ‘北京’

Took 0.0123 seconds.

# 插入多版本数据
hbase(main):022:0> put ‘fgedu_user’, ‘user_00000001’, ‘info:name’, ‘fgedu01_new’, 1680940800000

Took 0.0123 seconds.

# 使用脚本批量插入
hbase(main):023:0> load ‘put_data.rb’

# put_data.rb内容
put ‘fgedu_user’, ‘user_00000002’, ‘info:name’, ‘fgedu02’
put ‘fgedu_user’, ‘user_00000002’, ‘info:age’, ’30’
put ‘fgedu_user’, ‘user_00000002’, ‘info:city’, ‘上海’
put ‘fgedu_user’, ‘user_00000003’, ‘info:name’, ‘fgedu03’
put ‘fgedu_user’, ‘user_00000003’, ‘info:age’, ’28’
put ‘fgedu_user’, ‘user_00000003’, ‘info:city’, ‘广州’

# 使用ImportTsv批量导入
$ hbase org.apache.hadoop.hbase.mapreduce.ImportTsv \
-Dimporttsv.separator=’,’ \
-Dimporttsv.columns=HBASE_ROW_KEY,info:name,info:age,info:city \
fgedu_user \
/data/user_import.csv

# 使用Bulk Load导入
$ hbase org.apache.hadoop.hbase.mapreduce.LoadIncrementalHFiles \
/output/hfiles \
fgedu_user

3.2.2 数据查询

# 获取单行数据
hbase(main):024:0> get ‘fgedu_user’, ‘user_00000001’

COLUMN CELL
info:age timestamp=1680940800001, value=25
info:city timestamp=1680940800002, value=北京
info:name timestamp=1680940800000, value=fgedu01
1 row(s)
Took 0.0123 seconds.

# 获取指定列族
hbase(main):025:0> get ‘fgedu_user’, ‘user_00000001’, {COLUMN => ‘info’}

COLUMN CELL
info:age timestamp=1680940800001, value=25
info:city timestamp=1680940800002, value=北京
info:name timestamp=1680940800000, value=fgedu01
1 row(s)
Took 0.0123 seconds.

# 获取指定列
hbase(main):026:0> get ‘fgedu_user’, ‘user_00000001’, {COLUMN => ‘info:name’}

COLUMN CELL
info:name timestamp=1680940800000, value=fgedu01
1 row(s)
Took 0.0123 seconds.

# 获取多个版本
hbase(main):027:0> get ‘fgedu_user’, ‘user_00000001’, {COLUMN => ‘info:name’, VERSIONS => 3}

COLUMN CELL
info:name timestamp=1680940800000, value=fgedu01_new
info:name timestamp=1680940700000, value=fgedu01
2 row(s)
Took 0.0123 seconds.

# 获取指定时间范围
hbase(main):028:0> get ‘fgedu_user’, ‘user_00000001’, {TIMERANGE => [1680940700000, 1680940900000]}

COLUMN CELL
info:age timestamp=1680940800001, value=25
info:city timestamp=1680940800002, value=北京
info:name timestamp=1680940800000, value=fgedu01_new
1 row(s)
Took 0.0123 seconds.

# 扫描全表
hbase(main):029:0> scan ‘fgedu_user’

ROW COLUMN+CELL
user_00000001 column=info:age, timestamp=1680940800001, value=25
user_00000001 column=info:city, timestamp=1680940800002, value=北京
user_00000001 column=info:name, timestamp=1680940800000, value=fgedu01
user_00000002 column=info:age, timestamp=1680940800011, value=30
user_00000002 column=info:city, timestamp=1680940800012, value=上海
user_00000002 column=info:name, timestamp=1680940800010, value=fgedu02
user_00000003 column=info:age, timestamp=1680940800021, value=28
user_00000003 column=info:city, timestamp=1680940800022, value=广州
user_00000003 column=info:name, timestamp=1680940800020, value=fgedu03
3 row(s)
Took 0.0456 seconds.

# 范围扫描
hbase(main):030:0> scan ‘fgedu_user’, {STARTROW => ‘user_00000001’, STOPROW => ‘user_00000003’}

# 限制返回行数
hbase(main):031:0> scan ‘fgedu_user’, {LIMIT => 2}

# 过滤器查询
hbase(main):032:0> scan ‘fgedu_user’, {FILTER => “ValueFilter(=, ‘binary:25’)”}

ROW COLUMN+CELL
user_00000001 column=info:age, timestamp=1680940800001, value=25
1 row(s)
Took 0.0123 seconds.

# 统计行数
hbase(main):033:0> count ‘fgedu_user’

3 row(s)
Took 0.1234 seconds.
=> 3

3.2.3 数据删除

# 删除指定列
hbase(main):034:0> delete ‘fgedu_user’, ‘user_00000001’, ‘info:city’

Took 0.0123 seconds.

# 删除指定版本
hbase(main):035:0> delete ‘fgedu_user’, ‘user_00000001’, ‘info:name’, 1680940800000

Took 0.0123 seconds.

# 删除整行
hbase(main):036:0> deleteall ‘fgedu_user’, ‘user_00000003’

Took 0.0123 seconds.

# 删除指定列族
hbase(main):037:0> deleteall ‘fgedu_user’, ‘user_00000002’, ‘info’

Took 0.0123 seconds.

# 增量操作
hbase(main):038:0> increment ‘fgedu_user’, ‘user_00000001’, ‘info:login_count’, 1

COUNTER VALUE = 1
Took 0.0123 seconds.

# 追加操作
hbase(main):039:0> append ‘fgedu_user’, ‘user_00000001’, ‘info:tags’, ‘,vip’

CURRENT VALUE = vip,active,vip
Took 0.0123 seconds.

风哥提示：HBase Shell操作需要谨慎，特别是删除操作。建议在生产环境操作前先备份数据。更多学习教程公众号风哥教程itpux_com

3.3 管理操作实战

# 查看集群状态
hbase(main):040:0> status

active master: fgedu-node1:16000
1 backup masters
fgedu-node2:16000
3 live servers
fgedu-node3:16020
requestsPerSecond=10.0, numberOfOnlineRegions=10, usedHeapMB=1024, maxHeapMB=32768
fgedu-node4:16020
requestsPerSecond=8.0, numberOfOnlineRegions=10, usedHeapMB=1024, maxHeapMB=32768
fgedu-node5:16020
requestsPerSecond=12.0, numberOfOnlineRegions=10, usedHeapMB=1024, maxHeapMB=32768
0 dead servers
Master is fgedu-node1:16000

# 查看详细状态
hbase(main):041:0> status ‘detailed’

# 查看Region分布
hbase(main):042:0> list_regions ‘fgedu_user’

# 手动触发压缩
hbase(main):043:0> major_compact ‘fgedu_user’

Took 0.0123 seconds.

# 手动刷写MemStore
hbase(main):044:0> flush ‘fgedu_user’

Took 0.0123 seconds.

# 分裂Region
hbase(main):045:0> split ‘fgedu_user,,1680940800000.xxx’

Took 0.0123 seconds.

# 移动Region
hbase(main):046:0> move ‘region_encoded_name’, ‘fgedu-node3,16020,1680940800000’

Took 0.0123 seconds.

# 触发负载均衡
hbase(main):047:0> balancer

true
Took 0.0123 seconds.

# 开启/关闭负载均衡
hbase(main):048:0> balance_switch true

true
Took 0.0123 seconds.

# 创建快照
hbase(main):049:0> snapshot ‘fgedu_user’, ‘fgedu_user_snapshot_20260408’

Took 0.1234 seconds.

# 列出快照
hbase(main):050:0> list_snapshots

SNAPSHOT TABLE + CREATION TIME
fgedu_user_snapshot_20260408 fgedu_user (Mon Apr 08 13:00:00 +0800 2026)
1 row(s)
Took 0.0123 seconds.

# 恢复快照
hbase(main):051:0> disable ‘fgedu_user’

Took 0.5678 seconds.

hbase(main):052:0> restore_snapshot ‘fgedu_user_snapshot_20260408’

Took 0.7890 seconds.

hbase(main):053:0> enable ‘fgedu_user’

Took 0.5678 seconds.

# 删除快照
hbase(main):054:0> delete_snapshot ‘fgedu_user_snapshot_20260408’

Took 0.0123 seconds.

Part04-生产案例与实战讲解

4.1 用户表操作案例

# 用户表完整操作案例

# 1. 创建用户表
hbase(main):055:0> create ‘fgedu_user_profile’,
{NAME => ‘basic’, VERSIONS => 1, COMPRESSION => ‘SNAPPY’, BLOOMFILTER => ‘ROW’},
{NAME => ‘behavior’, VERSIONS => 3, COMPRESSION => ‘SNAPPY’, TTL => 2592000},
{NAME => ‘tag’, VERSIONS => 1, COMPRESSION => ‘SNAPPY’},
SPLITS => [‘10000000’, ‘20000000’, ‘30000000’, ‘40000000’, ‘50000000’, ‘60000000’, ‘70000000’, ‘80000000’, ‘90000000’]

Created table fgedu_user_profile

# 2. 插入用户数据
hbase(main):056:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘basic:name’, ‘fgedu01’
hbase(main):057:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘basic:age’, ’25’
hbase(main):058:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘basic:city’, ‘北京’
hbase(main):059:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘basic:phone’, ‘13800138001’
hbase(main):060:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘behavior:login_count’, ‘100’
hbase(main):061:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘behavior:buy_count’, ’10’
hbase(main):062:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘tag:level’, ‘VIP’
hbase(main):063:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘tag:interest’, ‘科技,旅游’

# 3. 查询用户数据
hbase(main):064:0> get ‘fgedu_user_profile’, ‘user_00000001’

COLUMN CELL
basic:age timestamp=1680940800001, value=25
basic:city timestamp=1680940800002, value=北京
basic:name timestamp=1680940800000, value=fgedu01
basic:phone timestamp=1680940800003, value=13800138001
behavior:buy_count timestamp=1680940800005, value=10
behavior:login_count timestamp=1680940800004, value=100
tag:interest timestamp=1680940800007, value=科技,旅游
tag:level timestamp=1680940800006, value=VIP
1 row(s)

# 4. 更新用户数据
hbase(main):065:0> put ‘fgedu_user_profile’, ‘user_00000001’, ‘basic:age’, ’26’
hbase(main):066:0> increment ‘fgedu_user_profile’, ‘user_00000001’, ‘behavior:login_count’, 1

# 5. 查看用户行为历史
hbase(main):067:0> get ‘fgedu_user_profile’, ‘user_00000001’, {COLUMN => ‘behavior’, VERSIONS => 3}

4.2 批量操作案例

# 批量操作脚本

# 创建批量插入脚本
$ cat > /tmp/batch_put.rb << 'EOF' # batch_put.rb # from:www.itpux.com.qq113257174.wx:itpux-com # web: http://www.fgedu.net.cn table = 'fgedu_user_profile' (1..1000).each do |i| user_id = format('user_%08d', i) name = "fgedu#{format('%02d', i % 100)}" age = (20 + rand(40)).to_s cities = ['北京', '上海', '广州', '深圳', '杭州'] city = cities[rand(cities.length)] put table, user_id, 'basic:name', name put table, user_id, 'basic:age', age put table, user_id, 'basic:city', city end puts "Batch insert completed!" EOF # 执行批量脚本 hbase(main):068:0> load ‘/tmp/batch_put.rb’

Batch insert completed!

# 使用ImportTsv导入数据
$ cat > /tmp/user_data.csv << 'EOF' user_00001001,fgedu1001,25,北京 user_00001002,fgedu1002,30,上海 user_00001003,fgedu1003,28,广州 EOF $ hbase org.apache.hadoop.hbase.mapreduce.ImportTsv \ -Dimporttsv.separator=',' \ -Dimporttsv.columns=HBASE_ROW_KEY,basic:name,basic:age,basic:city \ fgedu_user_profile \ /tmp/user_data.csv # 导出数据 $ hbase org.apache.hadoop.hbase.mapreduce.Export \ fgedu_user_profile \ /output/fgedu_user_profile_export

4.3 常见问题处理

4.3.1 命令执行慢

# 问题现象：命令执行缓慢

# 排查步骤
# 1. 检查RegionServer状态
hbase(main):069:0> status ‘detailed’

# 2. 检查Region分布
hbase(main):070:0> list_regions ‘fgedu_user_profile’

# 解决方案
# 1. 触发压缩
hbase(main):071:0> major_compact ‘fgedu_user_profile’

# 2. 检查BlockCache
# Web UI -> RegionServer -> Memory

# 3. 优化查询
# 使用过滤器减少返回数据
hbase(main):072:0> scan ‘fgedu_user_profile’,
{FILTER => “PrefixFilter(‘user_00001’)”, LIMIT => 100}

4.3.2 表被锁定

# 问题现象：表操作被锁定

# 排查步骤
# 1. 检查表状态
hbase(main):073:0> describe ‘fgedu_user_profile’

# 2. 检查Region状态
hbase(main):074:0> list_regions ‘fgedu_user_profile’

# 解决方案
# 1. 强制解锁
hbase(main):075:0> disable ‘fgedu_user_profile’
hbase(main):076:0> enable ‘fgedu_user_profile’

# 2. 检查是否有长时间运行的操作
# Web UI -> Procedures

Part05-风哥经验总结与分享

5.1 Shell最佳实践

HBase Shell最佳实践建议：

# Shell最佳实践
1. 使用脚本文件执行批量操作
2. 合理使用过滤器减少数据传输
3. 避免全表扫描
4. 定期执行压缩
5. 使用快照备份重要数据

5.2 操作建议

操作建议：

HBase Shell操作建议：

生产环境谨慎使用删除操作
大批量操作使用ImportTsv
定期创建快照备份
监控集群状态

5.3 工具推荐

HBase管理工具：

HBase Shell：命令行工具
HBase Web UI：Web管理界面
Phoenix：SQL层
Hue：Web界面

风哥提示：HBase Shell是HBase管理的重要工具，熟练掌握常用命令可以提高运维效率。建议使用脚本文件管理复杂操作。from bigdata视频:www.itpux.com

本文由风哥教程整理发布,仅用于学习测试使用,转载注明出处:http://www.fgedu.net.cn/10327.html

大数据教程FG074-HBase Shell命令操作实战

Part01-基础概念与理论知识

1.1 HBase Shell概述

1.2 命令分类

1.3 语法规则

Part02-生产环境规划与建议

2.1 表设计规划

2.2 RowKey设计规划

2.3 列族设计规划

Part03-生产环境项目实施方案

3.1 DDL操作实战

3.1.1 创建表

3.1.2 修改表

3.2 DML操作实战

3.2.1 数据插入

3.2.2 数据查询

3.2.3 数据删除

3.3 管理操作实战

Part04-生产案例与实战讲解

4.1 用户表操作案例

4.2 批量操作案例

4.3 常见问题处理

4.3.1 命令执行慢

4.3.2 表被锁定

Part05-风哥经验总结与分享

5.1 Shell最佳实践

5.2 操作建议

5.3 工具推荐

相关推荐

联系我们