Hbase介绍及操作
1. Hbase概述
1.1 Hbase是什么
- HBase是建立在HDFS之上的分布式面向列的数据库;属于KV结构数据,原生不支持标准SQL。它是一个Apache的开源项目,是横向扩展的。
- HBase可以提供快速随机访问海量结构化数据。它利用了Hadoop的文件系统(HDFS)提供的容错能力。
- HBase不同于一般的关系数据库,它是一个适合于非结构化数据存储的数据库,是HBase基于列的而不是基于行的模式。
1.2 Hbase数据单元
RowKey:是Byte array,是表中每条记录的“主键”,按照字典顺序排序,方便快速查找,Rowkey的设计非常重要;
Column Family:列族,拥有一个名称(string),包含一个或者多个相关列;
Column:属于某一个columnfamily,familyName:columnName,每条记录可动态添加;
Version Number:类型为Long,默认值是系统时间戳Timestamp,可由用户自定义;用于标记同一份数据的不同版本。
Value(Cell):Byte array;
2. Hbase的架构
2.1 Hbase的架构组成
HBase采用Master/Slave架构搭建集群,它隶属于Hadoop生态系统,由以下类型节点组成:HMaster节点、HRegionServer节点、ZooKeeper集群,而在底层,它将数据存储于HDFS中,因而涉及到HDFS的NameNode、DataNode等,总体结构如下:
各组件说明:
Client
- 使用HBase RPC机制与HMaster和HRegionServer进行通信;
- Client与HMaster进行通信进行管理类操作;
- Client与HRegionServer进行数据读写类操作;
HMaster
- HMaster没有单点问题,HBase中可以启动多个HMaster,通过Zookeeper保证总有一个Master在运行。
HMaster主要负责Table和Region的管理工作
- 管理用户对表的增删改查操作;
- 管理HRegionServer的负载均衡,调整Region分布;
- Region Split后,负责新Region的分布;
- 在HRegionServer停机后,负责失效HRegionServer上Region 的迁移;
HRegionServer
HBase中最核心的模块;
- 维护region,处理对这些region的IO请求;
- Regionserver负责切分在运行过程中变得过大的region;
HRegion
- HBase使用RowKey将表水平切割成多个HRegion,从HMaster的角度,每个HRegion都纪录了它的StartKey和EndKey(第一个HRegion的StartKey为空,最后一个HRegion的EndKey为空),由于RowKey是排序的,因而Client可以通过HMaster快速的定位每个RowKey在哪个HRegion中。
2.2查看hbase的web界面
3.hbase shell操作
3.1 DDL操作
#开启hbase shell
hbase shell
#查看hbase状态
status
#查看hbase版本
version
#创建命名空间
create_namespace '命名空间名'
#显示所有命名空间
list_namespace
#删除命名空间, 在删除一个命名空间时,该命名空间不能包含任何的表,否则会报错
drop_namespace '命名空间名'
#创建默认命名空间的表
create '表名称', '列族名称1','列族名称2','列族名称N'
#创建带有命名空间的表
create '命名空间:表名称', '列族名称1','列族名称2','列族名称N'
#列出所有表
list
#获得表的描述
describe '表名'
#删除table 表的 列族名称1 列族
alter 'table',{NAME=>'列族名称1',METHOD=>'delete'}
#删除多个列族
alter 'table', {NAME => '列族名称1', METHOD => 'delete'},{NAME => '列族名称2', METHOD => 'delete'}
#先把表下线
disable '表名'
#再drop表
drop '表名'
3.2 DML操作
#添加数据
# 语法:put <table>,<rowkey>,<family:column>,<value>,[<timestamp>]
#如果不写timestamp,则系统默认
put 'table','id01', 'c_f1:name','111'
#获取数据
#get: 获取表中一行数据,不能扫描全表
# 语法:get <table>,<rowkey>,[<family:column>,....]
get 'table','id01'
#更新数据
#语法:重新put,put时会覆盖原来的数据
put 'table','id01', 'c_f1:name','222'
#scan扫描
# 语法:scan <table> ,{COLUMNS => [ <family:column>,.... ], LIMIT => num}
#扫描全表,大表操作不可取
scan 'table'
#获取表中前两行
scan 'table', {LIMIT => 2}
#扫描表中指定列族数据
scan 'table', {COLUMNS => 'c_f1'}
#扫描表中执行列族中列的数据
scan 'table', {COLUMNS => 'c_f2:cert_no'}
#扫描表中值=222 的数据
scan 'table', FILTER=>"ValueFilter(=,'name:222')"
# 筛选行,按照rowkey的范围[STARTROW,STOPROW)
scan 'table', {STARTROW =>'id01' , STOPROW => 'id03'}
#删除行中某列数据
# 语法:delete <table>, <rowkey>, <family:column>
# 必须指定列名
# 会删除执行列的所有版本数据
delete 'table', 'id04', 'c_f2:name'
#删除整行
# 语法:deleteall <table>, <rowkey>
deleteall 'table', 'id05'
#清空表数据
# 语法: truncate <table>
truncate 'table'
#查询表中有多少行
# 语法:count <table>, {INTERVAL => intervalNum, CACHE => cacheNum}
# INTERVAL设置多少行显示一次及对应的rowkey,默认1000;
# CACHE每次去取的缓存区大小,默认是10,调整该参数可提高查询速度
#查询表中数据行数
count 'table'
#按照2行显示一次,查询
count 'table', {INTERVAL => 2}
4.hbase整合springboot
4.1pom
<dependency>
<groupId>org.apache.hbase</groupId>
<artifactId>hbase-client</artifactId>
<version>1.3.6</version>
</dependency>
4.2 application.properties
hbase.conf.confMaps.'hbase.zookeeper.quorum'=master,slave1,slave2
4.3 HbaseConfig 自定义配置类
import org.springframework.boot.context.properties.ConfigurationProperties;
import org.springframework.context.annotation.Configuration;
import java.util.Map;
/**
* Hbase-Conf配置
*
* @Author: suyuan
*/
@Configuration
@ConfigurationProperties(prefix = HbaseConfig.CONF_PREFIX)
public class HbaseConfig {
public static final String CONF_PREFIX = "hbase.conf";
private Map<String,String> confMaps;
public Map<String, String> getconfMaps() {
return confMaps;
}
public void setconfMaps(Map<String, String> confMaps) {
this.confMaps = confMaps;
}
}
4.4 HBaseUtils工具类
import org.springframework.beans.BeansException;
import org.springframework.context.ApplicationContext;
import org.springframework.context.ApplicationContextAware;
import org.springframework.stereotype.Component;
/**
* Spring的ApplicationContext的持有者,可以用静态方法的方式获取spring容器中的bean
*/
@Component
public class SpringContextHolder implements ApplicationContextAware {
private static ApplicationContext applicationContext;
@Override
public void setApplicationContext(ApplicationContext applicationContext) throws BeansException {
SpringContextHolder.applicationContext = applicationContext;
}
public static ApplicationContext getApplicationContext() {
assertApplicationContext();
return applicationContext;
}
@SuppressWarnings("unchecked")
public static <T> T getBean(String beanName) {
assertApplicationContext();
return (T) applicationContext.getBean(beanName);
}
public static <T> T getBean(Class<T> requiredType) {
assertApplicationContext();
return applicationContext.getBean(requiredType);
}
private static void assertApplicationContext() {
if (SpringContextHolder.applicationContext == null) {
throw new RuntimeException("applicaitonContext属性为null,请检查是否注入了SpringContextHolder!");
}
}
}
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.hbase.*;
import org.apache.hadoop.hbase.client.*;
import org.apache.hadoop.hbase.client.coprocessor.AggregationClient;
import org.apache.hadoop.hbase.client.coprocessor.LongColumnInterpreter;
import org.apache.hadoop.hbase.filter.*;
import org.apache.hadoop.hbase.util.Bytes;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.context.annotation.DependsOn;
import org.springframework.stereotype.Component;
import org.springframework.util.StopWatch;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;
import java.util.Map;
import java.util.NavigableMap;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
@DependsOn("springContextHolder") //控制依赖顺序,保证springContextHolder类在之前已经加载
@Component
public class HBaseUtils {
private Logger logger = LoggerFactory.getLogger(this.getClass());
//手动获取hbaseConfig配置类对象
private static HbaseConfig hbaseConfig = SpringContextHolder.getBean("hbaseConfig");
private static Configuration conf = HBaseConfiguration.create();
private static ExecutorService pool = Executors.newScheduledThreadPool(20); //设置连接池
private static Connection connection = null;
private static HBaseUtils instance = null;
private static Admin admin = null;
private HBaseUtils(){
if(connection == null){
try {
//将hbase配置类中定义的配置加载到连接池中每个连接里
Map<String, String> confMap = hbaseConfig.getconfMaps();
for (Map.Entry<String,String> confEntry : confMap.entrySet()) {
conf.set(confEntry.getKey(), confEntry.getValue());
}
connection = ConnectionFactory.createConnection(conf, pool);
admin = connection.getAdmin();
} catch (IOException e) {
logger.error("HbaseUtils实例初始化失败!错误信息为:" + e.getMessage(), e);
}
}
}
//简单单例方法,如果autowired自动注入就不需要此方法
public static synchronized HBaseUtils getInstance(){
if(instance == null){
instance = new HBaseUtils();
}
return instance;
}
/**
* 创建表
*
* @param tableName 表名
* @param columnFamily 列族(数组)
*/
public void createTable(String tableName, String[] columnFamily) throws IOException{
TableName name = TableName.valueOf(tableName);
//如果存在则删除
if (admin.tableExists(name)) {
admin.disableTable(name);
admin.deleteTable(name);
logger.error("create htable error! this table {} already exists!", name);
} else {
HTableDescriptor desc = new HTableDescriptor(name);
for (String cf : columnFamily) {
desc.addFamily(new HColumnDescriptor(cf));
}
admin.createTable(desc);
}
}
/**
* 插入记录(单行单列族-多列多值)
*
* @param tableName 表名
* @param row 行名
* @param columnFamilys 列族名
* @param columns 列名(数组)
* @param values 值(数组)(且需要和列一一对应)
*/
public void insertRecords(String tableName, String row, String columnFamilys, String[] columns, String[] values) throws IOException {
TableName name = TableName.valueOf(tableName);
Table table = connection.getTable(name);
Put put = new Put(Bytes.toBytes(row));
for (int i = 0; i < columns.length; i++) {
put.addColumn(Bytes.toBytes(columnFamilys), Bytes.toBytes(columns[i]), Bytes.toBytes(values[i]));
table.put(put);
}
}
/**
* 插入记录(单行单列族-单列单值)
*
* @param tableName 表名
* @param row 行名
* @param columnFamily 列族名
* @param column 列名
* @param value 值
*/
public void insertOneRecord(String tableName, String row, String columnFamily, String column, String value) throws IOException {
TableName name = TableName.valueOf(tableName);
Table table = connection.getTable(name);
Put put = new Put(Bytes.toBytes(row));
put.addColumn(Bytes.toBytes(columnFamily), Bytes.toBytes(column), Bytes.toBytes(value));
table.put(put);
}
/**
* 删除一行记录
*
* @param tablename 表名
* @param rowkey 行名
*/
public void deleteRow(String tablename, String rowkey) throws IOException {
TableName name = TableName.valueOf(tablename);
Table table = connection.getTable(name);
Delete d = new Delete(rowkey.getBytes());
table.delete(d);
}
/**
* 删除单行单列族记录
* @param tablename 表名
* @param rowkey 行名
* @param columnFamily 列族名
*/
public void deleteColumnFamily(String tablename, String rowkey, String columnFamily) throws IOException {
TableName name = TableName.valueOf(tablename);
Table table = connection.getTable(name);
Delete d = new Delete(rowkey.getBytes()).deleteFamily(Bytes.toBytes(columnFamily));
table.delete(d);
}
/**
* 删除单行单列族单列记录
*
* @param tablename 表名
* @param rowkey 行名
* @param columnFamily 列族名
* @param column 列名
*/
public void deleteColumn(String tablename, String rowkey, String columnFamily, String column) throws IOException {
TableName name = TableName.valueOf(tablename);
Table table = connection.getTable(name);
Delete d = new Delete(rowkey.getBytes()).deleteColumn(Bytes.toBytes(columnFamily), Bytes.toBytes(column));
table.delete(d);
}
/**
* 查找一行记录
*
* @param tablename 表名
* @param rowKey 行名
*/
public static String selectRow(String tablename, String rowKey) throws IOException {
String record = "";
TableName name=TableName.valueOf(tablename);
Table table = connection.getTable(name);
Get g = new Get(rowKey.getBytes());
Result rs = table.get(g);
NavigableMap<byte[], NavigableMap<byte[], NavigableMap<Long, byte[]>>> map = rs.getMap();
for (Cell cell : rs.rawCells()) {
StringBuffer stringBuffer = new StringBuffer().append(Bytes.toString(cell.getRow())).append("\t")
.append(Bytes.toString(cell.getFamily())).append("\t")
.append(Bytes.toString(cell.getQualifier())).append("\t")
.append(Bytes.toString(cell.getValue())).append("\n");
String str = stringBuffer.toString();
record += str;
}
return record;
}
/**
* 查找单行单列族单列记录
*
* @param tablename 表名
* @param rowKey 行名
* @param columnFamily 列族名
* @param column 列名
* @return
*/
public static String selectValue(String tablename, String rowKey, String columnFamily, String column) throws IOException {
TableName name=TableName.valueOf(tablename);
Table table = connection.getTable(name);
Get g = new Get(rowKey.getBytes());
g.addColumn(Bytes.toBytes(columnFamily), Bytes.toBytes(column));
Result rs = table.get(g);
return Bytes.toString(rs.value());
}
/**
* 查询表中所有行(Scan方式)
*
* @param tablename
* @return
*/
public String scanAllRecord(String tablename) throws IOException {
String record = "";
TableName name=TableName.valueOf(tablename);
Table table = connection.getTable(name);
Scan scan = new Scan();
ResultScanner scanner = table.getScanner(scan);
try {
for(Result result : scanner){
for (Cell cell : result.rawCells()) {
StringBuffer stringBuffer = new StringBuffer().append(Bytes.toString(cell.getRow())).append("\t")
.append(Bytes.toString(cell.getFamily())).append("\t")
.append(Bytes.toString(cell.getQualifier())).append("\t")
.append(Bytes.toString(cell.getValue())).append("\n");
String str = stringBuffer.toString();
record += str;
}
}
} finally {
if (scanner != null) {
scanner.close();
}
}
return record;
}
/**
* 根据rowkey关键字查询报告记录
*
* @param tablename
* @param rowKeyword
* @return
*/
public List scanReportDataByRowKeyword(String tablename, String rowKeyword) throws IOException {
ArrayList<Object> list = new ArrayList<>();
Table table = connection.getTable(TableName.valueOf(tablename));
Scan scan = new Scan();
//添加行键过滤器,根据关键字匹配
RowFilter rowFilter = new RowFilter(CompareFilter.CompareOp.EQUAL, new SubstringComparator(rowKeyword));
scan.setFilter(rowFilter);
ResultScanner scanner = table.getScanner(scan);
try {
for (Result result : scanner) {
//TODO 此处根据业务来自定义实现
list.add(null);
}
} finally {
if (scanner != null) {
scanner.close();
}
}
return list;
}
/**
* 根据rowkey关键字和时间戳范围查询报告记录
*
* @param tablename
* @param rowKeyword
* @return
*/
public List scanReportDataByRowKeywordTimestamp(String tablename, String rowKeyword, Long minStamp, Long maxStamp) throws IOException {
ArrayList<Object> list = new ArrayList<>();
Table table = connection.getTable(TableName.valueOf(tablename));
Scan scan = new Scan();
//添加scan的时间范围
scan.setTimeRange(minStamp, maxStamp);
RowFilter rowFilter = new RowFilter(CompareFilter.CompareOp.EQUAL, new SubstringComparator(rowKeyword));
scan.setFilter(rowFilter);
ResultScanner scanner = table.getScanner(scan);
try {
for (Result result : scanner) {
//TODO 此处根据业务来自定义实现
list.add(null);
}
} finally {
if (scanner != null) {
scanner.close();
}
}
return list;
}
/**
* 删除表操作
*
* @param tablename
*/
public void deleteTable(String tablename) throws IOException {
TableName name=TableName.valueOf(tablename);
if(admin.tableExists(name)) {
admin.disableTable(name);
admin.deleteTable(name);
}
}
/**
* 利用协处理器进行全表count统计
*
* @param tablename
*/
public Long countRowsWithCoprocessor(String tablename) throws Throwable {
TableName name=TableName.valueOf(tablename);
HTableDescriptor descriptor = admin.getTableDescriptor(name);
String coprocessorClass = "org.apache.hadoop.hbase.coprocessor.AggregateImplementation";
if (! descriptor.hasCoprocessor(coprocessorClass)) {
admin.disableTable(name);
descriptor.addCoprocessor(coprocessorClass);
admin.modifyTable(name, descriptor);
admin.enableTable(name);
}
//计时
StopWatch stopWatch = new StopWatch();
stopWatch.start();
Scan scan = new Scan();
AggregationClient aggregationClient = new AggregationClient(conf);
Long count = aggregationClient.rowCount(name, new LongColumnInterpreter(), scan);
stopWatch.stop();
System.out.println("RowCount:" + count + ",全表count统计耗时:" + stopWatch.getTotalTimeMillis());
return count;
}
}
4.5 测试用例
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RestController;
@RestController
@RequestMapping("/hbase")
public class HbaseController {
@Autowired
private HBaseUtils hBaseUtils;
/**
* 创建表
* @author suyuan
* @date 2021/6/22 17:17
*/
@RequestMapping("/createTable")
public String createTable() throws Exception {
String[] arr = {"name","age"};
hBaseUtils.createTable("sy:syy",arr);
return "ok";
}
/**
* 删除表
* @author suyuan
* @date 2021/6/22 17:18
*/
@RequestMapping("/deleteTable")
public String deleteTable() throws Exception {
hBaseUtils.deleteTable("syy");
return "ok";
}
/**
* 插入记录(插入同列族数据就是更新)
* @author suyuan
* @date 2021/6/22 17:18
*/
@RequestMapping("/insertRecords")
public String insertRecords() throws Exception {
hBaseUtils.insertRecords("sy:syy","20210622","name", new String[]{"name1","name2"}, new String[]{"111","222"});
return "ok";
}
/**
* 查找一行记录
* @author suyuan
* @date 2021/6/22 17:18
*/
@RequestMapping("/selectRow")
public String selectRow() throws Exception {
String s = hBaseUtils.selectRow("sy:syy", "20210622");
System.out.println(s);
return s;
}
}
本地编写代码需配置:
host地址
192.168.10.160 master
192.168.10.161 slave1
192.168.10.162 slave2
集群需要启动
zookeeper、hadoop、hbase