美文网首页
Spring相关面试题:谈一谈你对事务的理解?

Spring相关面试题:谈一谈你对事务的理解?

作者: Java编程日记 | 来源:发表于2022-02-17 13:06 被阅读0次

事务详解

2. 事务管理

2.1 回顾事务

什么是事务?

在数据库开发中,一组业务逻辑操作,要么全部成功,要么全部失败。

事务有什么特定?ACID

原子性:整体,原子不可分割的。整个操作被看成一个整体,要么成功,要么失败。

一致性:数据,事务操作的前后数据一致。

隔离性:并发,两个事务之间并发访问情况。

持久性:结果,事务一旦提交,不能回滚。

隔离有什么问题?

脏读:一个事务读到了另一个事务没有提交的数据。

不可重复读:一个事务读到了另一个事务已有提交的数据(update)。

幻读:一个事务读到了另一个事务已有提交的数据(insert)。

隔离级别有那些?

读未提交:存在3个问题(脏读、不可重复读、幻读)

读已提交:解决1个问题(脏读),存在2个问题(不可重复读、幻读)

可重复读:解决2个问题(脏读、不可重复读)、存在1个问题(幻读)

串行化:解决3个问题(脏读、不可重复读、幻读)–单事务

2.2 事务详解

研究Spring事务,需要学习事务管理平台管理器: PlatformTransactionManager 

在平台管理器中,通过 事务的定义 获得事务,从而进行事务提交或回滚操作。

事务定义 TransactionDefinition 的详解:

传播行为:一个事务调用另一个事务,事务共享问题。

PROPAGATION_REQUIRED,required:支持当前事务,如果没有事务,创建一个新的。

A 有事务,B使用A的事务。(支持当前事务)

A没有事务,B创建新的。()

PROPAGATION_SUPPORTS,supports:支持当前事务,如果没有事务,以非事务执行。

A 有事务,B使用A的事务。(支持当前事务)

A没有事务,B以非事务执行。

PROPAGATION_MANDATORY,mandatory:支持当前事务,如果没有事务,抛异常

A 有事务,B使用A的事务。(支持当前事务)

A没有事务,B抛异常。

PROPAGATION_REQUIRES_NEW,requires_new:创建一个新事物,如果当前有事务,将挂起。

A 有事务,B创建新事务,同时挂起A事务。

A 没有事务,B创建新事务。

PROPAGATION_NOT_SUPPORTED, not_supported:不支持当前事务,以非事务执行,如果有挂起

A 有事务,B以非事务执行,同时挂起A事务。

A 没有事务,B以非事务执行。

PROPAGATION_NEVER, never:不支持当前事务,如果有抛异常。

A 有事务,B抛异常

A 没有事务,B以非事务执行。

PROPAGATION_NESTED, nested :嵌套事务,底层使用 savepoint 进行嵌套事务操作。

保存点允许回顾部分事务。

相关源码:

/*

* Copyright 2002-2015 the original author or authors.

*

* Licensed under the Apache License, Version 2.0 (the "License");

* you may not use this file except in compliance with the License.

* You may obtain a copy of the License at

*

*      http://www.apache.org/licenses/LICENSE-2.0

*

* Unless required by applicable law or agreed to in writing, software

* distributed under the License is distributed on an "AS IS" BASIS,

* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.

* See the License for the specific language governing permissions and

* limitations under the License.

*/packageorg.springframework.transaction;importjava.sql.Connection;/** * Interface that defines Spring-compliant transaction properties. * Based on the propagation behavior definitions analogous to EJB CMT attributes. * *

Note that isolation level and timeout settings will not get applied unless * an actual new transaction gets started. As only {@link#PROPAGATION_REQUIRED}, * {@link#PROPAGATION_REQUIRES_NEW} and {@link#PROPAGATION_NESTED} can cause * that, it usually doesn't make sense to specify those settings in other cases. * Furthermore, be aware that not all transaction managers will support those * advanced features and thus might throw corresponding exceptions when given * non-default values. * *

The {@link#isReadOnly() read-only flag} applies to any transaction context, * whether backed by an actual resource transaction or operating non-transactionally * at the resource level. In the latter case, the flag will only apply to managed * resources within the application, such as a Hibernate {@codeSession}. * *@authorJuergen Hoeller *@since08.05.2003 *@seePlatformTransactionManager#getTransaction(TransactionDefinition) *@seeorg.springframework.transaction.support.DefaultTransactionDefinition *@seeorg.springframework.transaction.interceptor.TransactionAttribute */publicinterfaceTransactionDefinition{/**

  * Support a current transaction; create a new one if none exists.

  * Analogous to the EJB transaction attribute of the same name.

  * <p>This is typically the default setting of a transaction definition,

  * and typically defines a transaction synchronization scope.

  */intPROPAGATION_REQUIRED =0;/**  * Support a current transaction; execute non-transactionally if none exists.  * Analogous to the EJB transaction attribute of the same name.  *

NOTE: For transaction managers with transaction synchronization,  * {@codePROPAGATION_SUPPORTS} is slightly different from no transaction  * at all, as it defines a transaction scope that synchronization might apply to.  * As a consequence, the same resources (a JDBC {@codeConnection}, a  * Hibernate {@codeSession}, etc) will be shared for the entire specified  * scope. Note that the exact behavior depends on the actual synchronization  * configuration of the transaction manager!  *

In general, use {@codePROPAGATION_SUPPORTS} with care! In particular, do  * not rely on {@codePROPAGATION_REQUIRED} or {@codePROPAGATION_REQUIRES_NEW}  * within a {@codePROPAGATION_SUPPORTS} scope (which may lead to  * synchronization conflicts at runtime). If such nesting is unavoidable, make sure  * to configure your transaction manager appropriately (typically switching to  * "synchronization on actual transaction").  *@seeorg.springframework.transaction.support.AbstractPlatformTransactionManager#setTransactionSynchronization  *@seeorg.springframework.transaction.support.AbstractPlatformTransactionManager#SYNCHRONIZATION_ON_ACTUAL_TRANSACTION  */intPROPAGATION_SUPPORTS =1;/**  * Support a current transaction; throw an exception if no current transaction  * exists. Analogous to the EJB transaction attribute of the same name.  *

Note that transaction synchronization within a {@codePROPAGATION_MANDATORY}  * scope will always be driven by the surrounding transaction.  */intPROPAGATION_MANDATORY =2;/**  * Create a new transaction, suspending the current transaction if one exists.  * Analogous to the EJB transaction attribute of the same name.  *

NOTE: Actual transaction suspension will not work out-of-the-box  * on all transaction managers. This in particular applies to  * {@linkorg.springframework.transaction.jta.JtaTransactionManager},  * which requires the {@codejavax.transaction.TransactionManager} to be  * made available it to it (which is server-specific in standard Java EE).  *

A {@codePROPAGATION_REQUIRES_NEW} scope always defines its own  * transaction synchronizations. Existing synchronizations will be suspended  * and resumed appropriately.  *@seeorg.springframework.transaction.jta.JtaTransactionManager#setTransactionManager  */intPROPAGATION_REQUIRES_NEW =3;/**  * Do not support a current transaction; rather always execute non-transactionally.  * Analogous to the EJB transaction attribute of the same name.  *

NOTE: Actual transaction suspension will not work out-of-the-box  * on all transaction managers. This in particular applies to  * {@linkorg.springframework.transaction.jta.JtaTransactionManager},  * which requires the {@codejavax.transaction.TransactionManager} to be  * made available it to it (which is server-specific in standard Java EE).  *

Note that transaction synchronization is not available within a  * {@codePROPAGATION_NOT_SUPPORTED} scope. Existing synchronizations  * will be suspended and resumed appropriately.  *@seeorg.springframework.transaction.jta.JtaTransactionManager#setTransactionManager  */intPROPAGATION_NOT_SUPPORTED =4;/**  * Do not support a current transaction; throw an exception if a current transaction  * exists. Analogous to the EJB transaction attribute of the same name.  *

Note that transaction synchronization is not available within a  * {@codePROPAGATION_NEVER} scope.  */intPROPAGATION_NEVER =5;/**  * Execute within a nested transaction if a current transaction exists,  * behave like {@link#PROPAGATION_REQUIRED} else. There is no analogous  * feature in EJB.  *

NOTE: Actual creation of a nested transaction will only work on  * specific transaction managers. Out of the box, this only applies to the JDBC  * {@linkorg.springframework.jdbc.datasource.DataSourceTransactionManager}  * when working on a JDBC 3.0 driver. Some JTA providers might support  * nested transactions as well.  *@seeorg.springframework.jdbc.datasource.DataSourceTransactionManager  */intPROPAGATION_NESTED =6;/**  * Use the default isolation level of the underlying datastore.  * All other levels correspond to the JDBC isolation levels.  *@seejava.sql.Connection  */intISOLATION_DEFAULT = -1;/**  * Indicates that dirty reads, non-repeatable reads and phantom reads  * can occur.  *

This level allows a row changed by one transaction to be read by another  * transaction before any changes in that row have been committed (a "dirty read").  * If any of the changes are rolled back, the second transaction will have  * retrieved an invalid row.  *@seejava.sql.Connection#TRANSACTION_READ_UNCOMMITTED  */intISOLATION_READ_UNCOMMITTED = Connection.TRANSACTION_READ_UNCOMMITTED;/**  * Indicates that dirty reads are prevented; non-repeatable reads and  * phantom reads can occur.  *

This level only prohibits a transaction from reading a row  * with uncommitted changes in it.  *@seejava.sql.Connection#TRANSACTION_READ_COMMITTED  */intISOLATION_READ_COMMITTED = Connection.TRANSACTION_READ_COMMITTED;/**  * Indicates that dirty reads and non-repeatable reads are prevented;  * phantom reads can occur.  *

This level prohibits a transaction from reading a row with uncommitted changes  * in it, and it also prohibits the situation where one transaction reads a row,  * a second transaction alters the row, and the first transaction re-reads the row,  * getting different values the second time (a "non-repeatable read").  *@seejava.sql.Connection#TRANSACTION_REPEATABLE_READ  */intISOLATION_REPEATABLE_READ = Connection.TRANSACTION_REPEATABLE_READ;/**  * Indicates that dirty reads, non-repeatable reads and phantom reads  * are prevented.  *

This level includes the prohibitions in {@link#ISOLATION_REPEATABLE_READ}  * and further prohibits the situation where one transaction reads all rows that  * satisfy a {@codeWHERE} condition, a second transaction inserts a row  * that satisfies that {@codeWHERE} condition, and the first transaction  * re-reads for the same condition, retrieving the additional "phantom" row  * in the second read.  *@seejava.sql.Connection#TRANSACTION_SERIALIZABLE  */intISOLATION_SERIALIZABLE = Connection.TRANSACTION_SERIALIZABLE;/**

  * Use the default timeout of the underlying transaction system,

  * or none if timeouts are not supported.

  */intTIMEOUT_DEFAULT = -1;/**  * Return the propagation behavior.  *

Must return one of the {@codePROPAGATION_XXX} constants  * defined on {@linkTransactionDefinition this interface}.  *@returnthe propagation behavior  *@see#PROPAGATION_REQUIRED  *@seeorg.springframework.transaction.support.TransactionSynchronizationManager#isActualTransactionActive()  */intgetPropagationBehavior();/**  * Return the isolation level.  *

Must return one of the {@codeISOLATION_XXX} constants  * defined on {@linkTransactionDefinition this interface}.  *

Only makes sense in combination with {@link#PROPAGATION_REQUIRED}  * or {@link#PROPAGATION_REQUIRES_NEW}.  *

Note that a transaction manager that does not support custom isolation levels  * will throw an exception when given any other level than {@link#ISOLATION_DEFAULT}.  *@returnthe isolation level  */intgetIsolationLevel();/**  * Return the transaction timeout.  *

Must return a number of seconds, or {@link#TIMEOUT_DEFAULT}.  *

Only makes sense in combination with {@link#PROPAGATION_REQUIRED}  * or {@link#PROPAGATION_REQUIRES_NEW}.  *

Note that a transaction manager that does not support timeouts will throw  * an exception when given any other timeout than {@link#TIMEOUT_DEFAULT}.  *@returnthe transaction timeout  */intgetTimeout();/**  * Return whether to optimize as a read-only transaction.  *

The read-only flag applies to any transaction context, whether  * backed by an actual resource transaction  * ({@link#PROPAGATION_REQUIRED}/{@link#PROPAGATION_REQUIRES_NEW}) or  * operating non-transactionally at the resource level  * ({@link#PROPAGATION_SUPPORTS}). In the latter case, the flag will  * only apply to managed resources within the application, such as a  * Hibernate {@codeSession}.  <<  *

This just serves as a hint for the actual transaction subsystem;  * it will not necessarily cause failure of write access attempts.  * A transaction manager which cannot interpret the read-only hint will  * not throw an exception when asked for a read-only transaction.  *@return{@codetrue} if the transaction is to be optimized as read-only  *@seeorg.springframework.transaction.support.TransactionSynchronization#beforeCommit(boolean)  *@seeorg.springframework.transaction.support.TransactionSynchronizationManager#isCurrentTransactionReadOnly()  */booleanisReadOnly();/**  * Return the name of this transaction. Can be {@codenull}.  *

This will be used as the transaction name to be shown in a  * transaction monitor, if applicable (for example, WebLogic's).  *

In case of Spring's declarative transactions, the exposed name will be  * the {@codefully-qualified class name + "." + method name} (by default).  *@returnthe name of this transaction  *@seeorg.springframework.transaction.interceptor.TransactionAspectSupport  *@seeorg.springframework.transaction.support.TransactionSynchronizationManager#getCurrentTransactionName()  */StringgetName();}

相关文章

网友评论

      本文标题:Spring相关面试题:谈一谈你对事务的理解?

      本文链接:https://www.haomeiwen.com/subject/oaxslrtx.html