Spring+MyBatis数据读写分离的实例详解

本文介绍了Spring Boot + MyBatis读写分离,有需要了解Spring+MyBatis读写分离的朋友可参考。想了解更多相关信息请持续关注我们应届毕业生考试网!

其最终实现功能:

1.默认更新操作都使用写数据源

2.读操作都使用slave数据源

3.特殊设置:可以指定要使用的数据源类型及名称(如果有名称,则会根据名称使用相应的数据源)

其实现原理如下:

1.通过Spring AOP对dao层接口进行拦截,并对需要指定数据源的接口在ThradLocal中设置其数据源类型及名称

2.通过MyBatsi的插件,对根据更新或者查询操作在ThreadLocal中设置数据源(dao层没有指定的情况下)

3.继承AbstractRoutingDataSource类。

在此直接写死使用HikariCP作为数据源

其实现步骤如下:

1.定义其数据源配置文件并进行解析为数据源

2.定义AbstractRoutingDataSource类及其它注解

3.定义Aop拦截

4.定义MyBatis插件

5.整合在一起

1.配置及解析类

其配置参数直接使用HikariCP的配置,其具体参数可以参考HikariCP。

在此使用yaml格式,名称为datasource.yaml,内容如下:

dds:
write:
jdbcUrl: jdbc:mysql://localhost:3306/order
password: liu123
username: root
maxPoolSize: 10
minIdle: 3
poolName: master
read:
- jdbcUrl: jdbc:mysql://localhost:3306/test
password: liu123
username: root
maxPoolSize: 10
minIdle: 3
poolName: slave1
- jdbcUrl: jdbc:mysql://localhost:3306/test2
password: liu123
username: root
maxPoolSize: 10
minIdle: 3
poolName: slave2
定义该配置所对应的Bean,名称为DBConfig,内容如下:
@Component
@ConfigurationProperties(locations = "classpath:datasource.yaml", prefix = "dds")
public class DBConfig {
private List read;
private HikariConfig write;
public List getRead() {
return read;
}
public void setRead(List read) {
this.read = read;
}
public HikariConfig getWrite() {
return write;
}
public void setWrite(HikariConfig write) {
this.write = write;
}
}
把配置转换为DataSource的工具类,名称:DataSourceUtil,内容如下:
import com.zaxxer.hikari.HikariConfig;
import com.zaxxer.hikari.HikariDataSource;
import javax.sql.DataSource;
import java.util.ArrayList;
import java.util.List;
public class DataSourceUtil {
public static DataSource getDataSource(HikariConfig config) {
return new HikariDataSource(config);
}
public static List getDataSource(List configs) {
List result = null;
if (configs != null && configs.size() > 0) {
result = new ArrayList<>(configs.size());
for (HikariConfig config : configs) {
result.add(getDataSource(config));
}
} else {
result = new ArrayList<>(0);
}
return result;
}
}
2.注解及动态数据源
定义注解@DataSource,其用于需要对个别方法指定其要使用的数据源(如某个读操作需要在master上执行,但另一读方法b需要在读数据源的具体一台上面执行)
@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.METHOD)
public @interface DataSource {
/**
* 类型,代表是使用读还是写
* @return
*/
DataSourceType type() default DataSourceType.WRITE;
/**
* 指定要使用的'DataSource的名称
* @return
*/
String name() default "";
}
定义数据源类型,分为两种:READ,WRITE,内容如下:
public enum DataSourceType {
READ, WRITE;
}
定义保存这此共享信息的类DynamicDataSourceHolder,在其中定义了两个ThreadLocal和一个map,holder用于保存当前线程的数据源类型(读或者写),pool用于保存数据源名称(如果指定),其内容如下:
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
public class DynamicDataSourceHolder {
private static final Map cache = new ConcurrentHashMap<>();
private static final ThreadLocal holder = new ThreadLocal<>();
private static final ThreadLocal pool = new ThreadLocal<>();
public static void putToCache(String key, DataSourceType dataSourceType) {
cache.put(key,dataSourceType);
}
public static DataSourceType getFromCach(String key) {
return cache.get(key);
}
public static void putDataSource(DataSourceType dataSourceType) {
holder.set(dataSourceType);
}
public static DataSourceType getDataSource() {
return holder.get();
}
public static void putPoolName(String name) {
if (name != null && name.length() > 0) {
pool.set(name);
}
}
public static String getPoolName() {
return pool.get();
}
public static void clearDataSource() {
holder.remove();
pool.remove();
}
}
动态数据源类为DynamicDataSoruce,其继承自AbstractRoutingDataSource,可以根据返回的key切换到相应的数据源,其内容如下:
import com.zaxxer.hikari.HikariDataSource;
import org.springframework.jdbc.datasource.lookup.AbstractRoutingDataSource;
import javax.sql.DataSource;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ThreadLocalRandom;
public class DynamicDataSource extends AbstractRoutingDataSource {
private DataSource writeDataSource;
private List readDataSource;
private int readDataSourceSize;
private Map dataSourceMapping = new ConcurrentHashMap<>();
@Override
public void afterPropertiesSet() {
if (this.writeDataSource == null) {
throw new IllegalArgumentException("Property 'writeDataSource' is required");
}
setDefaultTargetDataSource(writeDataSource);
Map targetDataSource = new HashMap<>();
targetDataSource.put(DataSourceType.WRITE.name(), writeDataSource);
String poolName = ((HikariDataSource)writeDataSource).getPoolName();
if (poolName != null && poolName.length() > 0) {
dataSourceMapping.put(poolName,DataSourceType.WRITE.name());
}
if (this.readDataSource == null) {
readDataSourceSize = 0;
} else {
for (int i = 0; i < readDataSource.size(); i++) {
targetDataSource.put(DataSourceType.READ.name() + i, readDataSource.get(i));
poolName = ((HikariDataSource)readDataSource.get(i)).getPoolName();
if (poolName != null && poolName.length() > 0) {
dataSourceMapping.put(poolName,DataSourceType.READ.name() + i);
}
}
readDataSourceSize = readDataSource.size();
}
setTargetDataSources(targetDataSource);
super.afterPropertiesSet();
}
@Override
protected Object determineCurrentLookupKey() {
DataSourceType dataSourceType = DynamicDataSourceHolder.getDataSource();
String dataSourceName = null;
if (dataSourceType == null ||dataSourceType == DataSourceType.WRITE || readDataSourceSize == 0) {
dataSourceName = DataSourceType.WRITE.name();
} else {
String poolName = DynamicDataSourceHolder.getPoolName();
if (poolName == null) {
int idx = ThreadLocalRandom.current().nextInt(0, readDataSourceSize);
dataSourceName = DataSourceType.READ.name() + idx;
} else {
dataSourceName = dataSourceMapping.get(poolName);
}
}
DynamicDataSourceHolder.clearDataSource();
return dataSourceName;
}
public void setWriteDataSource(DataSource writeDataSource) {
this.writeDataSource = writeDataSource;
}
public void setReadDataSource(List readDataSource) {
this.readDataSource = readDataSource;
}
}