SpringBoot中怎么利用MyBatis实现乐观锁和悲观锁

SpringBoot中怎么利用MyBatis实现乐观锁和悲观锁,相信很多没有经验的人对此束手无策,为此本文总结了问题出现的原因和解决方法,通过这篇文章希望你能解决这个问题。

死锁问题

当 A, B 两个账户同时向对方转账时,会出现如下情况:

T1      Lock A      Lock B              T2      Lock B (由于事务 2 已经 Lock A,等待)      Lock A (由于事务 1 已经 Lock B,等待)

由于两个事务都在等待对方释放锁,于是死锁产生了,解决方案:按照主键的大小来加锁,总是先锁主键较小或较大的那行数据。

建立数据表并插入数据(MySQL

create table account(  id   int auto_increment    primary key,  deposit decimal(10, 2) default 0.00 not null,  version int      default 0  not null);INSERT INTO vault.account (id, deposit, version) VALUES (1, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (2, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (3, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (4, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (5, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (6, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (7, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (8, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (9, 1000, 0);INSERT INTO vault.account (id, deposit, version) VALUES (10, 1000, 0);

时刻 事务 1 (A 向 B 转账) 事务 2 (B 向 A 转账)

Mapper 文件

悲观锁使用 select ... for update,乐观锁使用 version 字段。

<?xml version="1.0" encoding="UTF-8" ?><!DOCTYPE mapper PUBLIC "-//mybatis.org//DTD Mapper 3.0//EN"    "http://mybatis.org/dtd/mybatis-3-mapper.dtd" ><mapper namespace="com.cloud.demo.mapper.AccountMapper">  <select id="selectById" resultType="com.cloud.demo.model.Account">    select *    from account    where id = #{id}  </select>  <update id="updateDeposit" keyProperty="id" parameterType="com.cloud.demo.model.Account">    update account    set deposit=#{deposit},      version = version + 1    where id = #{id}     and version = #{version}  </update>  <select id="selectByIdForUpdate" resultType="com.cloud.demo.model.Account">    select *    from account    where id = #{id} for    update  </select>  <update id="updateDepositPessimistic" keyProperty="id" parameterType="com.cloud.demo.model.Account">    update account    set deposit=#{deposit}    where id = #{id}  </update>  <select id="getTotalDeposit" resultType="java.math.BigDecimal">    select sum(deposit) from account;  </select></mapper>

Mapper 接口

@Componentpublic interface AccountMapper {  Account selectById(int id);  Account selectByIdForUpdate(int id);  int updateDepositWithVersion(Account account);  void updateDeposit(Account account);  BigDecimal getTotalDeposit();}

Account POJO

@Datapublic class Account {  private int id;  private BigDecimal deposit;  private int version;}

AccountService

在 transferOptimistic 方法上有个自定义注解 @Retry,这个用来实现乐观锁失败后重试。

@Slf4j@Servicepublic class AccountService {  public enum Result{    SUCCESS,    DEPOSIT_NOT_ENOUGH,    FAILED,  }  @Resource  private AccountMapper accountMapper;  private BiPredicate<BigDecimal, BigDecimal> isDepositEnough = (deposit, value) -> deposit.compareTo(value) > 0;  /**   * 转账操作,悲观锁   *   * @param fromId 扣款账户   * @param toId  收款账户   * @param value 金额   */  @Transactional(isolation = Isolation.READ_COMMITTED)  public Result transferPessimistic(int fromId, int toId, BigDecimal value) {    Account from, to;    try {      // 先锁 id 较大的那行,避免死锁      if (fromId > toId) {        from = accountMapper.selectByIdForUpdate(fromId);        to = accountMapper.selectByIdForUpdate(toId);      } else {        to = accountMapper.selectByIdForUpdate(toId);        from = accountMapper.selectByIdForUpdate(fromId);      }    } catch (Exception e) {      log.error(e.getMessage());      TransactionAspectSupport.currentTransactionStatus().setRollbackOnly();      return Result.FAILED;    }    if (!isDepositEnough.test(from.getDeposit(), value)) {      TransactionAspectSupport.currentTransactionStatus().setRollbackOnly();      log.info(String.format("Account %d is not enough.", fromId));      return Result.DEPOSIT_NOT_ENOUGH;    }    from.setDeposit(from.getDeposit().subtract(value));    to.setDeposit(to.getDeposit().add(value));    accountMapper.updateDeposit(from);    accountMapper.updateDeposit(to);    return Result.SUCCESS;  }  /**   * 转账操作,乐观锁   * @param fromId 扣款账户   * @param toId  收款账户   * @param value 金额   */  @Retry  @Transactional(isolation = Isolation.REPEATABLE_READ)  public Result transferOptimistic(int fromId, int toId, BigDecimal value) {    Account from = accountMapper.selectById(fromId),        to = accountMapper.selectById(toId);    if (!isDepositEnough.test(from.getDeposit(), value)) {      TransactionAspectSupport.currentTransactionStatus().setRollbackOnly();      return Result.DEPOSIT_NOT_ENOUGH;    }    from.setDeposit(from.getDeposit().subtract(value));    to.setDeposit(to.getDeposit().add(value));    int r1, r2;    // 先锁 id 较大的那行,避免死锁    if (from.getId() > to.getId()) {      r1 = accountMapper.updateDepositWithVersion(from);      r2 = accountMapper.updateDepositWithVersion(to);    } else {      r2 = accountMapper.updateDepositWithVersion(to);      r1 = accountMapper.updateDepositWithVersion(from);    }    if (r1 < 1 || r2 < 1) {      // 失败,抛出重试异常,执行重试      throw new RetryException("Transfer failed, retry.");    } else {      return Result.SUCCESS;    }  }}

使用 Spring AOP 实现乐观锁失败后重试

自定义注解 Retry

@Retention(RetentionPolicy.RUNTIME)@Target(ElementType.METHOD)public @interface Retry {  int value() default 3; // 重试次数}

重试异常 RetryException

public class RetryException extends RuntimeException {  public RetryException(String message) {    super(message);  }}

重试的切面类

tryAgain 方法使用了 @Around 注解(表示环绕通知),可以决定目标方法在何时执行,或者不执行,以及自定义返回结果。这里首先通过 ProceedingJoinPoint.proceed() 方法执行目标方法,如果抛出了重试异常,那么重新执行直到满三次,三次都不成功则回滚并返回 FAILED。

@Slf4j@Aspect@Componentpublic class RetryAspect {  @Pointcut("@annotation(com.cloud.demo.annotation.Retry)")  public void retryPointcut() {  }  @Around("retryPointcut() && @annotation(retry)")  @Transactional(isolation = Isolation.READ_COMMITTED)  public Object tryAgain(ProceedingJoinPoint joinPoint, Retry retry) throws Throwable {    int count = 0;    do {      count++;      try {        return joinPoint.proceed();      } catch (RetryException e) {        if (count > retry.value()) {          log.error("Retry failed!");          TransactionAspectSupport.currentTransactionStatus().setRollbackOnly();          return AccountService.Result.FAILED;        }      }    } while (true);  }}

单元测试

用多个线程模拟并发转账,经过测试,悲观锁除了账户余额不足,或者数据库连接不够以及等待超时,全部成功;乐观锁即使加了重试,成功的线程也很少,500 个平均也就十几个成功。

所以对于写多读少的操作,使用悲观锁,对于读多写少的操作,可以使用乐观锁。

完整代码请见 Github:https://github.com/imcloudfloating/Lock_Demo。

@Slf4j@SpringBootTest@RunWith(SpringRunner.class)class AccountServiceTest {  // 并发数  private static final int COUNT = 500;  @Resource  AccountMapper accountMapper;  @Resource  AccountService accountService;  private CountDownLatch latch = new CountDownLatch(COUNT);  private List<Thread> transferThreads = new ArrayList<>();  private List<Pair<Integer, Integer>> transferAccounts = new ArrayList<>();  @BeforeEach  void setUp() {    Random random = new Random(currentTimeMillis());    transferThreads.clear();    transferAccounts.clear();    for (int i = 0; i < COUNT; i++) {      int from = random.nextInt(10) + 1;      int to;      do{        to = random.nextInt(10) + 1;      } while (from == to);      transferAccounts.add(new Pair<>(from, to));    }  }  /**   * 测试悲观锁   */  @Test  void transferByPessimisticLock() throws Throwable {    for (int i = 0; i < COUNT; i++) {      transferThreads.add(new Transfer(i, true));    }    for (Thread t : transferThreads) {      t.start();    }    latch.await();    Assertions.assertEquals(accountMapper.getTotalDeposit(),        BigDecimal.valueOf(10000).setScale(2, RoundingMode.HALF_UP));  }  /**   * 测试乐观锁   */  @Test  void transferByOptimisticLock() throws Throwable {    for (int i = 0; i < COUNT; i++) {      transferThreads.add(new Transfer(i, false));    }    for (Thread t : transferThreads) {      t.start();    }    latch.await();    Assertions.assertEquals(accountMapper.getTotalDeposit(),        BigDecimal.valueOf(10000).setScale(2, RoundingMode.HALF_UP));  }  /**   * 转账线程   */  class Transfer extends Thread {    int index;    boolean isPessimistic;    Transfer(int i, boolean b) {      index = i;      isPessimistic = b;    }    @Override    public void run() {      BigDecimal value = BigDecimal.valueOf(          new Random(currentTimeMillis()).nextFloat() * 100      ).setScale(2, RoundingMode.HALF_UP);      AccountService.Result result = AccountService.Result.FAILED;      int fromId = transferAccounts.get(index).getKey(),          toId = transferAccounts.get(index).getValue();      try {        if (isPessimistic) {          result = accountService.transferPessimistic(fromId, toId, value);        } else {          result = accountService.transferOptimistic(fromId, toId, value);        }      } catch (Exception e) {        log.error(e.getMessage());      } finally {        if (result == AccountService.Result.SUCCESS) {          log.info(String.format("Transfer %f from %d to %d success", value, fromId, toId));        }        latch.countDown();      }    }  }}

MySQL 配置

innodb_rollback_on_timeout='ON'max_connections=1000innodb_lock_wait_timeout=500

看完上述内容,你们掌握SpringBoot中怎么利用MyBatis实现乐观锁和悲观锁的方法了吗?如果还想学到更多技能或想了解更多相关内容,欢迎关注行业资讯频道,感谢各位的阅读!