在Spring中，事务有两种实现方式：
- 编程式事务管理： 编程式事务管理使用TransactionTemplate可实现更细粒度的事务控制。
- 申明式事务管理： 基于Spring AOP实现。其本质是对方法前后进行拦截，然后在目标方法开始之前创建或者加入一个事务，在执行完目标方法之后根据执行情况提交或者回滚事务。
申明式事务管理不需要入侵代码，通过@Transactional就可以进行事务操作，更快捷而且简单（尤其是配合spring boot自动配置），且大部分业务都可以满足，推荐使用。其实不管是编程式事务还是申明式事务，最终调用的底层核心代码是一致的。在进入主题之前，我们最好是对 Spring AOP的织入及调用过程、以及 Spring的事务传播基本知识 需要有一点的了解。直接进入正题
编程式事务 TransactionTemplate ：
对于编程式事务来说，其核心是 execute 方法
TransactionTemplate transactionTemplate = new TransactionTemplate();
transactionTemplate.execute(new TransactionCallback<Object>() {
@Override
public Object doInTransaction(TransactionStatus status) {
return null;
}
});
先来看一下其类图：
看到了两个接口 TransactionOperations, InitializingBean。其中后者就不多介绍了，对于前者，定义了一个执行方法，并且在TransactionTemplate进行了实现：
public interface TransactionOperations {
/**
* Execute the action specified by the given callback object within a transaction.
* <p>Allows for returning a result object created within the transaction, that is,
* a domain object or a collection of domain objects. A RuntimeException thrown
* by the callback is treated as a fatal exception that enforces a rollback.
* Such an exception gets propagated to the caller of the template.
* @param action the callback object that specifies the transactional action
* @return a result object returned by the callback, or {@code null} if none
* @throws TransactionException in case of initialization, rollback, or system errors
* @throws RuntimeException if thrown by the TransactionCallback
*/
@Nullable
<T> T execute(TransactionCallback<T> action) throws TransactionException;
}
接下去看这两个接口在 TransactionTemplate 中的实现：
@Override
public void afterPropertiesSet() {
//仅仅做了以下判断，校验事务管理器是否为空
if (this.transactionManager == null) {
throw new IllegalArgumentException("Property 'transactionManager' is required");
}
}
//这里面的操作很清晰
@Override
@Nullable
public <T> T execute(TransactionCallback<T> action) throws TransactionException {
Assert.state(this.transactionManager != null, "No PlatformTransactionManager set");
// 判断类型
if (this.transactionManager instanceof CallbackPreferringPlatformTransactionManager) {
return ((CallbackPreferringPlatformTransactionManager) this.transactionManager).execute(this, action);
}
else {//开启一个事务
TransactionStatus status = this.transactionManager.getTransaction(this);
T result;
try {//调用我们的业务实现逻辑
result = action.doInTransaction(status);
}
catch (RuntimeException | Error ex) {
// Transactional code threw application exception -> rollback
// 异常就回滚事务
rollbackOnException(status, ex);
throw ex;
}
catch (Throwable ex) {
// Transactional code threw unexpected exception -> rollback
// 异常就回滚事务
rollbackOnException(status, ex);
throw new UndeclaredThrowableException(ex, "TransactionCallback threw undeclared checked exception");
}
// 提交事务
this.transactionManager.commit(status);
return result;
}
}
具体对应的实现在声明式事务分析到这些操作的源码的时候会进一步分析，都是调用 PlatformTransactionManager相对应的方法。
声明式事务 @Transactional：
在springboot中我们要开启事务需要在主启动类上引入 @EnableTransactionManagement 注解。然后对需要的方法上打上@Transactional，在实际开发中我们很少在类上直接声明 @Transactional，在类上注解即代表该类所有的 piblic 的方法都需要开启事务。还有一个知识点是Springboot中的事务拓展了 @EnableAutoConfiguration 来实现自动装配，所以在spring-boot-autoconfigure包里面可以看到会导入两个类
重点来看 org.springframework.boot.autoconfigure.transaction.TransactionAutoConfiguration：
@Configuration
@ConditionalOnClass(PlatformTransactionManager.class)//类路径下包含PlatformTransactionManager这个类时这个自动配置生效
@AutoConfigureAfter({ JtaAutoConfiguration.class, HibernateJpaAutoConfiguration.class,
DataSourceTransactionManagerAutoConfiguration.class,
Neo4jDataAutoConfiguration.class })//在这几个类注入后在开始载入
@EnableConfigurationProperties(TransactionProperties.class)
public class TransactionAutoConfiguration {
@Bean
@ConditionalOnMissingBean
public TransactionManagerCustomizers platformTransactionManagerCustomizers(
ObjectProvider<List<PlatformTransactionManagerCustomizer<?>>> customizers) {
return new TransactionManagerCustomizers(customizers.getIfAvailable());
}
//事务模板配置类
@Configuration
@ConditionalOnSingleCandidate(PlatformTransactionManager.class)
public static class TransactionTemplateConfiguration {
private final PlatformTransactionManager transactionManager;
public TransactionTemplateConfiguration(
PlatformTransactionManager transactionManager) {
this.transactionManager = transactionManager;
}
@Bean
@ConditionalOnMissingBean//如果没有定义TransactionTemplate bean生成一个
public TransactionTemplate transactionTemplate() {
return new TransactionTemplate(this.transactionManager);
}
}
@Configuration
@ConditionalOnBean(PlatformTransactionManager.class)
//Configuration支持2种代理方式
@ConditionalOnMissingBean(AbstractTransactionManagementConfiguration.class)
public static class EnableTransactionManagementConfiguration {
//即proxyTargetClass = false表示是JDK动态代理支持的是：面向接口代理。
@Configuration
@EnableTransactionManagement(proxyTargetClass = false)
@ConditionalOnProperty(prefix = "spring.aop", name = "proxy-target-class", havingValue = "false", matchIfMissing = false)
public static class JdkDynamicAutoProxyConfiguration {
//即spring.aop.proxy-target-class=false时生效，且没有这个配置不生效。
}
//即proxyTargetClass = true标识Cglib代理支持的是子类继承代理。
@Configuration
@EnableTransactionManagement(proxyTargetClass = true)
@ConditionalOnProperty(prefix = "spring.aop", name = "proxy-target-class", havingValue = "true", matchIfMissing = true)
public static class CglibAutoProxyConfiguration {
//即spring.aop.proxy-target-class=true时生效，且没有这个配置默认生效。默认没有配置默认为 true，走的Cglib代理。说明@Transactional注解支持直接加在类上。
}
}
}
紧接着我们就需要来看看 @EnableTransactionManagement 这个注解了，也正是我们在启动类上打上的注解：
@Target(ElementType.TYPE)
@Retention(RetentionPolicy.RUNTIME)
@Documented
@Import(TransactionManagementConfigurationSelector.class)
public @interface EnableTransactionManagement {
boolean proxyTargetClass() default false;
AdviceMode mode() default AdviceMode.PROXY;
int order() default Ordered.LOWEST_PRECEDENCE;
}
引入了 TransactionManagementConfigurationSelector.class
public class TransactionManagementConfigurationSelector extends AdviceModeImportSelector<EnableTransactionManagement> {
@Override
protected String[] selectImports(AdviceMode adviceMode) {
switch (adviceMode) {
case PROXY: // 默认走这里，因为上面哪个注解中默认就是 PROXY
return new String[] {AutoProxyRegistrar.class.getName(), ProxyTransactionManagementConfiguration.class.getName()};
case ASPECTJ:
return new String[] {TransactionManagementConfigUtils.TRANSACTION_ASPECT_CONFIGURATION_CLASS_NAME};
default:
return null;
}
}
}
在该类中引入了 AutoProxyRegistrar.class, ProxyTransactionManagementConfiguration.class 两个类
- AutoProxyRegistrar：向容器注入一个 key 为 org.springframework.aop.config.internalAutoProxyCreator ，value 为 InfrastructureAdvisorAutoProxyCreator.class 的 bean定义。
ProxyTransactionManagementConfiguration：代理事务配置类，定义事务拦截器相关类。
注入的这个过程就不详细介绍了，很常规的操作。然后先来看 InfrastructureAdvisorAutoProxyCreator：
重点关注右上角的那两个类：
- BeanPostProcessor：定义了2个方法，作用于类初始化的前后处理逻辑。
- InstantiationAwareBeanPostProcessor：定义了3个方法，作用于实例化的前后处理逻辑。
该4个前后处理器的方法实现在 AbstractAutoProxyCreator 中进行了实现：
//实例化前置处理器
@Override
public Object postProcessBeforeInstantiation(Class<?> beanClass, String beanName) throws BeansException {
Object cacheKey = getCacheKey(beanClass, beanName);
if (!StringUtils.hasLength(beanName) || !this.targetSourcedBeans.contains(beanName)) {
if (this.advisedBeans.containsKey(cacheKey)) {
return null;
}//是否是这四类实现：Advice、Pointcut、Advisor、AopInfrastructureBean
if (isInfrastructureClass(beanClass) || shouldSkip(beanClass, beanName)) {
//标记是否需要增强实现，供后面postProcessAfterInitialization实例化后使用。
this.advisedBeans.put(cacheKey, Boolean.FALSE);
return null;
}
}
//如果我们有一个自定义的TargetSource，在这里创建代理。很显然目前来说没有定义
// Create proxy here if we have a custom TargetSource.
// Suppresses unnecessary default instantiation of the target bean:
// The TargetSource will handle target instances in a custom fashion.
TargetSource targetSource = getCustomTargetSource(beanClass, beanName);
if (targetSource != null) {
if (StringUtils.hasLength(beanName)) {
this.targetSourcedBeans.add(beanName);
}
Object[] specificInterceptors = getAdvicesAndAdvisorsForBean(beanClass, beanName, targetSource);
Object proxy = createProxy(beanClass, beanName, specificInterceptors, targetSource);
this.proxyTypes.put(cacheKey, proxy.getClass());
return proxy;
}
return null;
}
//实例化后置处理器
@Override
public boolean postProcessAfterInstantiation(Object bean, String beanName) {
return true;
}
// 初始化前置处理器
@Override
public Object postProcessBeforeInitialization(Object bean, String beanName) {
return bean;
}
/**初始化后置处理器，就是在AOP中提到的
* Create a proxy with the configured interceptors if the bean is
* identified as one to proxy by the subclass.
* @see #getAdvicesAndAdvisorsForBean
*/
@Override
public Object postProcessAfterInitialization(@Nullable Object bean, String beanName) throws BeansException {
if (bean != null) {
Object cacheKey = getCacheKey(bean.getClass(), beanName);
if (!this.earlyProxyReferences.contains(cacheKey)) {
return wrapIfNecessary(bean, beanName, cacheKey);
}
}
return bean;
}
然后再初始化后置处理方法中的 wrapIfNecessary(增强织入)：
protected Object wrapIfNecessary(Object bean, String beanName, Object cacheKey) {
// 如果是用户自定义获取实例，不需要增强处理，直接返回
if (beanName != null && this.targetSourcedBeans.contains(beanName)) {
return bean;
}// 查询map缓存，标记过false,不需要增强直接返回
if (Boolean.FALSE.equals(this.advisedBeans.get(cacheKey))) {
return bean;
}// 判断一遍springAOP基础构建类，标记过false,不需要增强直接返回
if (isInfrastructureClass(bean.getClass()) || shouldSkip(bean.getClass(), beanName)) {
this.advisedBeans.put(cacheKey, Boolean.FALSE);
return bean;
}
// 获取增强List<Advisor> advisors 这里就会获取到一个 TransationIntercepter的advisors
Object[] specificInterceptors = getAdvicesAndAdvisorsForBean(bean.getClass(), beanName, null);
// 如果存在增强
if (specificInterceptors != DO_NOT_PROXY) {
this.advisedBeans.put(cacheKey, Boolean.TRUE);// 标记增强为TRUE,表示需要增强实现
// 生成增强代理类
Object proxy = createProxy(
bean.getClass(), beanName, specificInterceptors, new SingletonTargetSource(bean));
this.proxyTypes.put(cacheKey, proxy.getClass());
return proxy;
}
// 如果不存在增强，标记false,作为缓存，再次进入提高效率，第16行利用缓存先校验
this.advisedBeans.put(cacheKey, Boolean.FALSE);
return bean;
}
然后进入创建代理类，与AOP中探讨的 JDK动态代理不一样的地方如下：
- 这里默认会走 CGLib 动态代理。
- 这里形成的调用链会走 getCallbacks 所获取到的 拦截器链。
- 会调用 DynamicAdvisedInterceptor 的intercept 方法创建CglibMethodInvocation 进入 process方法
ProxyTransactionManagementConfiguration：
先来看一下其源码中都定义了什么：
@Configuration
public class ProxyTransactionManagementConfiguration extends AbstractTransactionManagementConfiguration {
// 这个bean会在后置处理器织入AOP的时候被扫描到
@Bean(name = TransactionManagementConfigUtils.TRANSACTION_ADVISOR_BEAN_NAME)
@Role(BeanDefinition.ROLE_INFRASTRUCTURE)//定义事务增强
public BeanFactoryTransactionAttributeSourceAdvisor transactionAdvisor() {
BeanFactoryTransactionAttributeSourceAdvisor advisor = new BeanFactoryTransactionAttributeSourceAdvisor();
advisor.setTransactionAttributeSource(transactionAttributeSource());
advisor.setAdvice(transactionInterceptor());
if (this.enableTx != null) {
advisor.setOrder(this.enableTx.<Integer>getNumber("order"));
}
return advisor;
}
// 基于注解的事务资源信息
@Bean
@Role(BeanDefinition.ROLE_INFRASTRUCTURE)
public TransactionAttributeSource transactionAttributeSource() {
return new AnnotationTransactionAttributeSource();
}
// 定义事务拦截器
@Bean
@Role(BeanDefinition.ROLE_INFRASTRUCTURE)
public TransactionInterceptor transactionInterceptor() {
TransactionInterceptor interceptor = new TransactionInterceptor();
interceptor.setTransactionAttributeSource(transactionAttributeSource());
if (this.txManager != null) {
interceptor.setTransactionManager(this.txManager);
}
return interceptor;
}
}
TransactionInterceptor 事务处理流程：
TransactionInterceptor使用通用的spring事务基础架构实现“声明式事务”，继承自TransactionAspectSupport类（该类包含与Spring的底层事务API的集成），实现了MethodInterceptor接口。spring类图如下：
事务拦截器的拦截功能就是依靠实现了MethodInterceptor接口，这个是spring的方法拦截器，并在invoke方法中调用了父类TransactionAspectSupport的invokeWithinTransaction()方法，源码如下(全体注意！关键的地方来了)：
public Object invoke(final MethodInvocation invocation) throws Throwable { // Work out the target class: may be {@code null}. // The TransactionAttributeSource should be passed the target class // as well as the method, which may be from an interface. Class<?> targetClass = (invocation.getThis() != null ? AopUtils.getTargetClass(invocation.getThis()) : null); // Adapt to TransactionAspectSupport's invokeWithinTransaction... return invokeWithinTransaction(invocation.getMethod(), targetClass, invocation::proceed); } protected Object invokeWithinTransaction(Method method, @Nullable Class<?> targetClass, final InvocationCallback invocation) throws Throwable { // 这里指的 non-transaction 是指编程式事务 // If the transaction attribute is null, the method is non-transactional. TransactionAttributeSource tas = getTransactionAttributeSource(); final TransactionAttribute txAttr = (tas != null ? tas.getTransactionAttribute(method, targetClass) : null); final PlatformTransactionManager tm = determineTransactionManager(txAttr); final String joinpointIdentification = methodIdentification(method, targetClass, txAttr); //声明式事务：事务属性为空、或者事务管理器非这个类型 if (txAttr == null || !(tm instanceof CallbackPreferringPlatformTransactionManager)) { // Standard transaction demarcation with getTransaction and commit/rollback calls. // 如果需要创建事务 TransactionInfo txInfo = createTransactionIfNecessary(tm, txAttr, joinpointIdentification); Object retVal = null; try { // This is an around advice: Invoke the next interceptor in the chain. // This will normally result in a target object being invoked. //方法执行、环绕增强、可自定义 retVal = invocation.proceedWithInvocation(); } catch (Throwable ex) { // target invocation exception // 事务异常，提交/回滚 completeTransactionAfterThrowing(txInfo, ex); throw ex; } finally { // 清除事务信息、重置为老的事务 cleanupTransactionInfo(txInfo); } // 提交事务 commitTransactionAfterReturning(txInfo); return retVal; } // 很明显是编程式事务 else { final ThrowableHolder throwableHolder = new ThrowableHolder(); // It's a CallbackPreferringPlatformTransactionManager: pass a TransactionCallback in. try { Object result = ((CallbackPreferringPlatformTransactionManager) tm).execute(txAttr, status -> { // 准备事务信息 TransactionInfo txInfo = prepareTransactionInfo(tm, txAttr, joinpointIdentification, status); try {//环绕增强 return invocation.proceedWithInvocation(); } catch (Throwable ex) { // 回滚 if (txAttr.rollbackOn(ex)) { // A RuntimeException: will lead to a rollback. if (ex instanceof RuntimeException) { throw (RuntimeException) ex; } else { throw new ThrowableHolderException(ex); } } else { // A normal return value: will lead to a commit. throwableHolder.throwable = ex; return null; } } finally {//清除事务 cleanupTransactionInfo(txInfo); } }); // Check result state: It might indicate a Throwable to rethrow. if (throwableHolder.throwable != null) { throw throwableHolder.throwable; } return result; } // ...... } }
其中 invocation::proceed 是lamda表达式，对应的是：
new InvocationCallback() { public Object proceedWithInvocation() throws Throwable { return invocation.proceed(); }
createTransactionIfNecessary 创建事务：
protected TransactionInfo createTransactionIfNecessary(@Nullable PlatformTransactionManager tm,
@Nullable TransactionAttribute txAttr, final String joinpointIdentification) {
// If no name specified, apply method identification as transaction name.
// ....省略代码
TransactionStatus status = null;
if (txAttr != null) {
if (tm != null) {
// 根据事务属性获得status
status = tm.getTransaction(txAttr);
}
else {
if (logger.isDebugEnabled()) {
logger.debug("Skipping transactional joinpoint [" + joinpointIdentification +
"] because no transaction manager has been configured");
}
}
}
return prepareTransactionInfo(tm, txAttr, joinpointIdentification, status);
}
//通过事务管理器 PlatformTransactionManager 的 getTransaction开启事务
public final TransactionStatus getTransaction(@Nullable TransactionDefinition definition) throws TransactionException {
// .....省略代码
try {
boolean newSynchronization = (getTransactionSynchronization() != SYNCHRONIZATION_NEVER);
DefaultTransactionStatus status = newTransactionStatus(
definition, transaction, true, newSynchronization, debugEnabled, suspendedResources);
doBegin(transaction, definition); // 开启事务 connection.setAutoCommit(false)
prepareSynchronization(status, definition);
return status;
}
// .....省略代码
}
//构造TransactionInfo 对象绑定当前线程
protected TransactionInfo prepareTransactionInfo(@Nullable PlatformTransactionManager tm,
@Nullable TransactionAttribute txAttr, String joinpointIdentification,
@Nullable TransactionStatus status) {
TransactionInfo txInfo = new TransactionInfo(tm, txAttr, joinpointIdentification);
// .....省略代码
txInfo.bindToThread();
return txInfo;
}
invocation.proceed()回调业务方法，这个 invication 也就是刚刚获取代理的时候生成了一个回调方法集合。其中有个 DynamicAdvisedInterceptor 拦截器。最终会走到他的 intercept 方法：
public Object intercept(Object proxy, Method method, Object[] args, MethodProxy methodProxy) throws Throwable {
Object oldProxy = null;
boolean setProxyContext = false;
Object target = null;
TargetSource targetSource = this.advised.getTargetSource();
try {
if (this.advised.exposeProxy) {
// Make invocation available if necessary.
oldProxy = AopContext.setCurrentProxy(proxy);
setProxyContext = true;
}
// Get as late as possible to minimize the time we "own" the target, in case it comes from a pool...
target = targetSource.getTarget();
Class<?> targetClass = (target != null ? target.getClass() : null);
//获取 advice组成chain
List<Object> chain = this.advised.getInterceptorsAndDynamicInterceptionAdvice(method, targetClass);
Object retVal;
// Check whether we only have one InvokerInterceptor: that is,
// no real advice, but just reflective invocation of the target.
if (chain.isEmpty() && Modifier.isPublic(method.getModifiers())) {
// We can skip creating a MethodInvocation: just invoke the target directly.
// Note that the final invoker must be an InvokerInterceptor, so we know
// it does nothing but a reflective operation on the target, and no hot
// swapping or fancy proxying.
Object[] argsToUse = AopProxyUtils.adaptArgumentsIfNecessary(method, args);
retVal = methodProxy.invoke(target, argsToUse);
}
else {
// We need to create a method invocation...
retVal = new CglibMethodInvocation(proxy, target, method, args, targetClass, chain, methodProxy).proceed();
}
retVal = processReturnType(proxy, target, method, retVal);
return retVal;
}
finally {
if (target != null && !targetSource.isStatic()) {
targetSource.releaseTarget(target);
}
if (setProxyContext) {
// Restore old proxy.
AopContext.setCurrentProxy(oldProxy);
}
}
}
大家可以看到这里构造了 CglibMethodInvocation 这么个对象，并且调用了他的 process 方法。我们先来看一下他的类图：
原来他是 ReflectiveMethodInvocation 的子类。ReflectiveMethodInvocation正式JDK动态代理中所生成的 MethodInvocation。进入他的 proceed()。这个代码对于看过AOP源码的朋友来说可以说是很熟悉了。根据组装的拦截器链进行链式处理。最后循环回来等链处理完了，反射调用原来的方法。
@Override
public Object proceed() throws Throwable {
// 启动时索引为-1，唤醒连接点，后续递增
if (this.currentInterceptorIndex == this.interceptorsAndDynamicMethodMatchers.size() - 1) {
return invokeJoinpoint(); // 这里回调真正的业务逻辑
}
Object interceptorOrInterceptionAdvice =
this.interceptorsAndDynamicMethodMatchers.get(++this.currentInterceptorIndex);
if (interceptorOrInterceptionAdvice instanceof InterceptorAndDynamicMethodMatcher) {
// 这里进行动态方法匹配校验，静态的方法匹配早已经校验过了（MethodMatcher接口有两种典型：动态/静态校验）
InterceptorAndDynamicMethodMatcher dm =
(InterceptorAndDynamicMethodMatcher) interceptorOrInterceptionAdvice;
if (dm.methodMatcher.matches(this.method, this.targetClass, this.arguments)) {
return dm.interceptor.invoke(this);
}
else {
// 动态匹配失败，跳过当前拦截，进入下一个（拦截器链）
return proceed();
}
}
else {
// 它是一个拦截器，所以我们只调用它:在构造这个对象之前，切入点将被静态地计算。
return ((MethodInterceptor) interceptorOrInterceptionAdvice).invoke(this);
}
}
朋友们可以调试一下代码，这里的走的 ((MethodInterceptor) interceptorOrInterceptionAdvice).invoke(this)逻辑， 在我们这个场景下走的就是 TransactionInterceptor 的 invoke方法的入口。这里需要理解的是第一次进入 process() 走else调用了TransactionInterceptor 的 invoke方法，然后invocation.proceedWithInvocation() 进入process方法的时候调用链已经走完了，那么这个时候就直接走真正的业务逻辑 invokeJoinpoint()。这就是spring事务的基本织入及调用过程。
PlatformTransactionManager 事务管理器：
无论在声明式事务、编程式事务中，最后都会走事务管理器的PlatformTransactionManager接口的3个方法：
public interface PlatformTransactionManager {
TransactionStatus getTransaction(@Nullable TransactionDefinition definition) throws TransactionException;
void commit(TransactionStatus status) throws TransactionException;
void rollback(TransactionStatus status) throws TransactionException;
}
我们先来看一下常见的实现类图：
如上图所示，PlatformTransactionManager顶级接口定义了最核心的事务管理方法，下面一层是AbstractPlatformTransactionManager抽象类，实现了PlatformTransactionManager接口的方法并定义了一些抽象方法，供子类拓展。最后下面一层是2个经典事务管理器：
- DataSourceTransactionmanager,即JDBC单数据库事务管理器，基于Connection实现，也是我们这次讨论的重点。
- JtaTransactionManager,即多数据库事务管理器（又叫做分布式事务管理器），其实现了JTA规范，使用XA协议进行两阶段提交。
我们这里只看基于DataSourceTransactionmanager源码。
getTransaction 事务获取：
@Override public final TransactionStatus getTransaction(TransactionDefinition definition) throws TransactionException { Object transaction = doGetTransaction(); // Cache debug flag to avoid repeated checks. boolean debugEnabled = logger.isDebugEnabled(); if (definition == null) { // Use defaults if no transaction definition given. definition = new DefaultTransactionDefinition(); } // 如果当前已经存在事务 if (isExistingTransaction(transaction)) { // 根据不同传播机制不同处理、就不走下面了 return handleExistingTransaction(definition, transaction, debugEnabled); } // 超时不能小于默认值 if (definition.getTimeout() < TransactionDefinition.TIMEOUT_DEFAULT) { throw new InvalidTimeoutException("Invalid transaction timeout", definition.getTimeout()); } // 当前不存在事务，传播机制=MANDATORY（支持当前事务，没事务可以让他支持），抛出异常 if (definition.getPropagationBehavior() == TransactionDefinition.PROPAGATION_MANDATORY) { throw new IllegalTransactionStateException( "No existing transaction found for transaction marked with propagation 'mandatory'"); }// 当前不存在事务，传播机制为 REQUIRED/REQUIRED_NEW/NESTED,这三种情况，需要新开启事务，且加上事务同步 else if (definition.getPropagationBehavior() == TransactionDefinition.PROPAGATION_REQUIRED || definition.getPropagationBehavior() == TransactionDefinition.PROPAGATION_REQUIRES_NEW || definition.getPropagationBehavior() == TransactionDefinition.PROPAGATION_NESTED) { SuspendedResourcesHolder suspendedResources = suspend(null); if (debugEnabled) { logger.debug("Creating new transaction with name [" + definition.getName() + "]: " + definition); } try {// 是否需要新开启同步// 开启// 开启 boolean newSynchronization = (getTransactionSynchronization() != SYNCHRONIZATION_NEVER); DefaultTransactionStatus status = newTransactionStatus( definition, transaction, true, newSynchronization, debugEnabled, suspendedResources); doBegin(transaction, definition);// 开启新事务 prepareSynchronization(status, definition);//预备同步 return status; } catch (RuntimeException ex) { resume(null, suspendedResources); throw ex; } catch (Error err) { resume(null, suspendedResources); throw err; } } else { // 当前不存在事务当前不存在事务，且传播机制=PROPAGATION_SUPPORTS/PROPAGATION_NOT_SUPPORTED/PROPAGATION_NEVER，
// 这三种情况，创建“空”事务:没有实际事务，但可能是同步。
// 警告：定义了隔离级别，但并没有真实的事务初始化，隔离级别被忽略有隔离级别但是并没有定义实际的事务初始化，有隔离级别但是并没有定义实际的事务初始化， if (definition.getIsolationLevel() != TransactionDefinition.ISOLATION_DEFAULT && logger.isWarnEnabled()) { logger.warn("Custom isolation level specified but no actual transaction initiated; " + "isolation level will effectively be ignored: " + definition); } boolean newSynchronization = (getTransactionSynchronization() == SYNCHRONIZATION_ALWAYS); return prepareTransactionStatus(definition, null, true, newSynchronization, debugEnabled, null); } }
这里的代码我们终于遇到了Spring的事务传播属性相关的东西，这里分成了两种情况：
- isExistingTransaction 判断是否已经存在事务，若存在进入handleExistingTransaction()根据不同传播机制不同处理（嵌套事务）
- isExistingTransaction 返回false，不存在已有事务，根据事务传播属性进行分类处理
先来看嵌套事务：
private TransactionStatus handleExistingTransaction(
TransactionDefinition definition, Object transaction, boolean debugEnabled)
throws TransactionException {
// 1.NERVER（不支持当前事务;如果当前事务存在，抛出异常）报错
if (definition.getPropagationBehavior() == TransactionDefinition.PROPAGATION_NEVER) {
throw new IllegalTransactionStateException(
"Existing transaction found for transaction marked with propagation 'never'");
}
// 2.NOT_SUPPORTED（不支持当前事务，现有同步将被挂起）挂起当前事务
if (definition.getPropagationBehavior() == TransactionDefinition.PROPAGATION_NOT_SUPPORTED) {
if (debugEnabled) {
logger.debug("Suspending current transaction");
}
Object suspendedResources = suspend(transaction);
boolean newSynchronization = (getTransactionSynchronization() == SYNCHRONIZATION_ALWAYS);
return prepareTransactionStatus(
definition, null, false, newSynchronization, debugEnabled, suspendedResources);
}
// 3.REQUIRES_NEW挂起当前事务，创建新事务
if (definition.getPropagationBehavior() == TransactionDefinition.PROPAGATION_REQUIRES_NEW) {
if (debugEnabled) {
logger.debug("Suspending current transaction, creating new transaction with name [" +
definition.getName() + "]");
}// 挂起当前事务
SuspendedResourcesHolder suspendedResources = suspend(transaction);
try {// 创建新事务
boolean newSynchronization = (getTransactionSynchronization() != SYNCHRONIZATION_NEVER);
DefaultTransactionStatus status = newTransactionStatus(
definition, transaction, true, newSynchronization, debugEnabled, suspendedResources);
doBegin(transaction, definition);
prepareSynchronization(status, definition);
return status;
}
catch (RuntimeException beginEx) {
resumeAfterBeginException(transaction, suspendedResources, beginEx);
throw beginEx;
}
catch (Error beginErr) {
resumeAfterBeginException(transaction, suspendedResources, beginErr);
throw beginErr;
}
}
// 4.NESTED嵌套事务
if (definition.getPropagationBehavior() == TransactionDefinition.PROPAGATION_NESTED) {
if (!isNestedTransactionAllowed()) {
throw new NestedTransactionNotSupportedException(
"Transaction manager does not allow nested transactions by default - " +
"specify 'nestedTransactionAllowed' property with value 'true'");
}
if (debugEnabled) {
logger.debug("Creating nested transaction with name [" + definition.getName() + "]");
}// 是否支持保存点：非JTA事务走这个分支。AbstractPlatformTransactionManager默认是true，JtaTransactionManager复写了该方法false，DataSourceTransactionmanager没有复写，还是true,
if (useSavepointForNestedTransaction()) {
// Usually uses JDBC 3.0 savepoints. Never activates Spring synchronization.
DefaultTransactionStatus status =
prepareTransactionStatus(definition, transaction, false, false, debugEnabled, null);
status.createAndHoldSavepoint();// 创建保存点
return status;
}
else {
// JTA事务走这个分支，创建新事务
boolean newSynchronization = (getTransactionSynchronization() != SYNCHRONIZATION_NEVER);
DefaultTransactionStatus status = newTransactionStatus(
definition, transaction, true, newSynchronization, debugEnabled, null);
doBegin(transaction, definition);
prepareSynchronization(status, definition);
return status;
}
}
// .....省略部分代码 // 到这里PROPAGATION_SUPPORTS 或 PROPAGATION_REQUIRED或PROPAGATION_MANDATORY，存在事务加入事务即可，prepareTransactionStatus第三个参数就是是否需要新事务。false代表不需要新事物
boolean newSynchronization = (getTransactionSynchronization() != SYNCHRONIZATION_NEVER);
return prepareTransactionStatus(definition, transaction, false, newSynchronization, debugEnabled, null);
}
上面的代码段已经反映出了嵌套事务的规则，不过其中又两个比较重要的方法：
- doBegin(transaction, definition) 开启一个新事务。
- suspend(transaction) 挂起当前事务
suspend(transaction):先来看挂起事务，其中先解除绑定当前事务各种属性：名称、只读、隔离级别、是否是真实的事务:
protected final SuspendedResourcesHolder suspend(Object transaction) throws TransactionException {
if (TransactionSynchronizationManager.isSynchronizationActive()) {// 1.当前存在同步，
List<TransactionSynchronization> suspendedSynchronizations = doSuspendSynchronization();
try {
Object suspendedResources = null;
if (transaction != null) {// 事务不为空，挂起事务
suspendedResources = doSuspend(transaction);
}// 解除绑定当前事务各种属性：名称、只读、隔离级别、是否是真实的事务.
String name = TransactionSynchronizationManager.getCurrentTransactionName();
TransactionSynchronizationManager.setCurrentTransactionName(null);
boolean readOnly = TransactionSynchronizationManager.isCurrentTransactionReadOnly();
TransactionSynchronizationManager.setCurrentTransactionReadOnly(false);
Integer isolationLevel = TransactionSynchronizationManager.getCurrentTransactionIsolationLevel();
TransactionSynchronizationManager.setCurrentTransactionIsolationLevel(null);
boolean wasActive = TransactionSynchronizationManager.isActualTransactionActive();
TransactionSynchronizationManager.setActualTransactionActive(false);
return new SuspendedResourcesHolder(
suspendedResources, suspendedSynchronizations, name, readOnly, isolationLevel, wasActive);
}
// ..... 省略代码
}
然后调用 doSuspend(transaction) ，这个方法由具体的实现类去复写，这里就是DataSourceTransactionManager：
@Override
protected Object doSuspend(Object transaction) {
DataSourceTransactionObject txObject = (DataSourceTransactionObject) transaction;
txObject.setConnectionHolder(null);//把当前事务的connectionHolder数据库连接持有者清空。
// 当前线程解绑datasource.其实就是ThreadLocal移除对应变量private static final ThreadLocal<Map<Object, Object>> resources = new NamedThreadLocal<>("Transactional resources");
return TransactionSynchronizationManager.unbindResource(obtainDataSource());
}
上述代码中就涉及到一个很重要的事务同步管理器 ：TransactionSynchronizationManager ，该类中维护了事务中本地变量：
public abstract class TransactionSynchronizationManager {
private static final Log logger = LogFactory.getLog(TransactionSynchronizationManager.class);
// 事务资源：map<k,v> 两种数据对。1.会话工厂和会话k=SqlsessionFactory v=SqlSessionHolder 2.数据源和连接k=DataSource v=ConnectionHolder，就是上面提到的map
private static final ThreadLocal<Map<Object, Object>> resources =
new NamedThreadLocal<Map<Object, Object>>("Transactional resources");
// 事务同步
private static final ThreadLocal<Set<TransactionSynchronization>> synchronizations =
new NamedThreadLocal<Set<TransactionSynchronization>>("Transaction synchronizations");
// 当前事务名称
private static final ThreadLocal<String> currentTransactionName =
new NamedThreadLocal<String>("Current transaction name");
// 当前事务的只读属性
private static final ThreadLocal<Boolean> currentTransactionReadOnly =
new NamedThreadLocal<Boolean>("Current transaction read-only status");
// 当前事务的隔离级别
private static final ThreadLocal<Integer> currentTransactionIsolationLevel =
new NamedThreadLocal<Integer>("Current transaction isolation level");
// 是否存在事务
private static final ThreadLocal<Boolean> actualTransactionActive =
new NamedThreadLocal<Boolean>("Actual transaction active");
// ....省略代码
}
doBegin(transaction, definition) :
@Override
protected void doBegin(Object transaction, TransactionDefinition definition) {
DataSourceTransactionObject txObject = (DataSourceTransactionObject) transaction;
Connection con = null;
try {// 如果事务还没有connection或者connection在事务同步状态，重置新的connectionHolder
if (!txObject.hasConnectionHolder() ||
txObject.getConnectionHolder().isSynchronizedWithTransaction()) {
Connection newCon = this.dataSource.getConnection();
if (logger.isDebugEnabled()) {
logger.debug("Acquired Connection [" + newCon + "] for JDBC transaction");
}// 重置新的connectionHolder
txObject.setConnectionHolder(new ConnectionHolder(newCon), true);
}
//设置新的连接为事务同步中
txObject.getConnectionHolder().setSynchronizedWithTransaction(true);
con = txObject.getConnectionHolder().getConnection();
//conn设置事务隔离级别,只读
Integer previousIsolationLevel = DataSourceUtils.prepareConnectionForTransaction(con, definition);
txObject.setPreviousIsolationLevel(previousIsolationLevel);//DataSourceTransactionObject设置事务隔离级别
// 如果是自动提交切换到手动提交
// so we don't want to do it unnecessarily (for example if we've explicitly
// configured the connection pool to set it already).
if (con.getAutoCommit()) {
txObject.setMustRestoreAutoCommit(true);
if (logger.isDebugEnabled()) {
logger.debug("Switching JDBC Connection [" + con + "] to manual commit");
}
con.setAutoCommit(false);
}
// 如果只读，执行sql设置事务只读
prepareTransactionalConnection(con, definition);
txObject.getConnectionHolder().setTransactionActive(true);// 设置connection持有者的事务开启状态
int timeout = determineTimeout(definition);
if (timeout != TransactionDefinition.TIMEOUT_DEFAULT) {
txObject.getConnectionHolder().setTimeoutInSeconds(timeout);// 设置超时秒数
}
// 绑定connection持有者到当前线程
if (txObject.isNewConnectionHolder()) {
TransactionSynchronizationManager.bindResource(getDataSource(), txObject.getConnectionHolder());
}
}
//....省略代码
}
开启新事务的 doBegin() 中最重要的操作：
- DataSourceTransactionObject“数据源事务对象”，设置ConnectionHolder，再给ConnectionHolder设置各种属性：自动提交、超时、事务开启、隔离级别。
- 给当前线程绑定一个线程本地变量，key=DataSource数据源 v=ConnectionHolder数据库连接。
commit 提交事务：
public final void commit(TransactionStatus status) throws TransactionException {
if (status.isCompleted()) {//事务已经完成，抛出异常
throw new IllegalTransactionStateException(
"Transaction is already completed - do not call commit or rollback more than once per transaction");
}
DefaultTransactionStatus defStatus = (DefaultTransactionStatus) status;
if (defStatus.isLocalRollbackOnly()) {//事务状态标记为回滚
if (defStatus.isDebug()) {
logger.debug("Transactional code has requested rollback");
}//执行回滚
processRollback(defStatus, false);
return;
}
if (!shouldCommitOnGlobalRollbackOnly() && defStatus.isGlobalRollbackOnly()) {
if (defStatus.isDebug()) {
logger.debug("Global transaction is marked as rollback-only but transactional code requested commit");
}//回滚
processRollback(defStatus, true);
return;
}
// 处理提交
processCommit(defStatus);
}
经过一系列判断进入主要的方法 processCommit:
private void processCommit(DefaultTransactionStatus status) throws TransactionException {
try {
boolean beforeCompletionInvoked = false;
try {//3个前置操作
prepareForCommit(status);
triggerBeforeCommit(status);
triggerBeforeCompletion(status);
beforeCompletionInvoked = true;//3个前置操作已调用
boolean globalRollbackOnly = false;//新事务 或 全局回滚失败
if (status.isNewTransaction() || isFailEarlyOnGlobalRollbackOnly()) {
globalRollbackOnly = status.isGlobalRollbackOnly();
}//1.有保存点，即嵌套事务
if (status.hasSavepoint()) {
if (status.isDebug()) {
logger.debug("Releasing transaction savepoint");
}//释放保存点
status.releaseHeldSavepoint();
}//2.新事务
else if (status.isNewTransaction()) {
if (status.isDebug()) {
logger.debug("Initiating transaction commit");
}//调用事务处理器提交事务
doCommit(status);
}
// 3.非新事务，且全局回滚失败，但是提交时没有得到异常，抛出异常
if (globalRollbackOnly) {
throw new UnexpectedRollbackException(
"Transaction silently rolled back because it has been marked as rollback-only");
}
}
catch (UnexpectedRollbackException ex) {
// 触发完成后事务同步，状态为回滚
triggerAfterCompletion(status, TransactionSynchronization.STATUS_ROLLED_BACK);
throw ex;
}// 事务异常
catch (TransactionException ex) {
// 提交失败回滚
if (isRollbackOnCommitFailure()) {
doRollbackOnCommitException(status, ex);
}// 触发完成后回调，事务同步状态为未知
else {
triggerAfterCompletion(status, TransactionSynchronization.STATUS_UNKNOWN);
}
throw ex;
}// 运行时异常
catch (RuntimeException ex) {
// 如果3个前置步骤未完成，调用前置的最后一步操作
if (!beforeCompletionInvoked) {
triggerBeforeCompletion(status);
}// 提交异常回滚
doRollbackOnCommitException(status, ex);
throw ex;
}// 其它异常
catch (Error err) {
// 如果3个前置步骤未完成，调用前置的最后一步操作
if (!beforeCompletionInvoked) {
triggerBeforeCompletion(status);
}// 提交异常回滚
doRollbackOnCommitException(status, err);
throw err;
}
// Trigger afterCommit callbacks, with an exception thrown there
// propagated to callers but the transaction still considered as committed.
try {
triggerAfterCommit(status);
}
finally {
triggerAfterCompletion(status, TransactionSynchronization.STATUS_COMMITTED);
}
}
finally {
cleanupAfterCompletion(status);
}
}
细节部分就不去扣了，这里最终调用 doCommit(status) 提交事务：
@Override protected void doCommit(DefaultTransactionStatus status) { DataSourceTransactionObject txObject = (DataSourceTransactionObject) status.getTransaction(); Connection con = txObject.getConnectionHolder().getConnection(); if (status.isDebug()) { logger.debug("Committing JDBC transaction on Connection [" + con + "]"); } try {//提交事务 con.commit(); } catch (SQLException ex) { throw new TransactionSystemException("Could not commit JDBC transaction", ex); } }
提交流程中涉及到的SqlSessionSynchronization是SqlSessionUtils的一个内部类，继承自TransactionSynchronizationAdapter抽象类，实现了事务同步接口TransactionSynchronization。看一下接口定义
public interface TransactionSynchronization extends Flushable {
/** 正确提交时的完成状态 */
int STATUS_COMMITTED = 0;
/** 正确回滚情况下的完成状态 */
int STATUS_ROLLED_BACK = 1;
/** 启发式混合完成或系统错误情况下的完成状态 */
int STATUS_UNKNOWN = 2;
/**挂起事务，解绑资源*/
default void suspend() {
}
/**唤醒事务，重新绑定资源*/
default void resume() {
}
/**将底层会话刷新到数据存储(如果适用)*/
@Override
default void flush() {
}
/**事务提交前的操作*/
default void beforeCommit(boolean readOnly) {
}
/**事务提交完成前的操作*/
default void beforeCompletion() {
}
/**事务提交后的额操作*/
default void afterCommit() {
}
/**事务提交完成后后的操作*/
default void afterCompletion(int status) {
}
}
rollback 回滚事务：
public final void rollback(TransactionStatus status) throws TransactionException {
if (status.isCompleted()) { // 事务状态是完成，抛出异常
throw new IllegalTransactionStateException(
"Transaction is already completed - do not call commit or rollback more than once per transaction");
}
DefaultTransactionStatus defStatus = (DefaultTransactionStatus) status;
processRollback(defStatus, false);//回滚
}
processRollback 回滚：
private void processRollback(DefaultTransactionStatus status) {
try {
try {// 解绑当前线程绑定的会话工厂，并关闭会话
triggerBeforeCompletion(status);
if (status.hasSavepoint()) {// 1.如果有保存点，即嵌套式事务
if (status.isDebug()) {
logger.debug("Rolling back transaction to savepoint");
}//回滚到保存点
status.rollbackToHeldSavepoint();
}//2.如果就是一个简单事务
else if (status.isNewTransaction()) {
if (status.isDebug()) {
logger.debug("Initiating transaction rollback");
}//回滚核心方法
doRollback(status);
}//3.当前存在事务且没有保存点，即加入当前事务的
else if (status.hasTransaction()) {//如果已经标记为回滚 或 当加入事务失败时全局回滚（默认true）
if (status.isLocalRollbackOnly() || isGlobalRollbackOnParticipationFailure()) {
if (status.isDebug()) {//debug时会打印：加入事务失败-标记已存在事务为回滚
logger.debug("Participating transaction failed - marking existing transaction as rollback-only");
}//设置当前connectionHolder：当加入一个已存在事务时回滚
doSetRollbackOnly(status);
}
else {
if (status.isDebug()) {
logger.debug("Participating transaction failed - letting transaction originator decide on rollback");
}
}
}
else {
logger.debug("Should roll back transaction but cannot - no transaction available");
}
}
catch (RuntimeException ex) {//关闭会话，重置SqlSessionHolder属性
triggerAfterCompletion(status, TransactionSynchronization.STATUS_UNKNOWN);
throw ex;
}
catch (Error err) {
triggerAfterCompletion(status, TransactionSynchronization.STATUS_UNKNOWN);
throw err;
}
triggerAfterCompletion(status, TransactionSynchronization.STATUS_ROLLED_BACK);
}
finally {//解绑当前线程
cleanupAfterCompletion(status);
}
}
调用回滚核心方法doRollback(status)：
protected void doRollback(DefaultTransactionStatus status) {
DataSourceTransactionObject txObject = (DataSourceTransactionObject) status.getTransaction();
Connection con = txObject.getConnectionHolder().getConnection();
if (status.isDebug()) {
logger.debug("Rolling back JDBC transaction on Connection [" + con + "]");
}
try {//回滚
con.rollback();
}
catch (SQLException ex) {
throw new TransactionSystemException("Could not roll back JDBC transaction", ex);
}
}
整个流程深度还是比较深的，需要我们慢慢去琢磨。慢慢积累吧！