Spring Optimistic Locking:How to retry transactional method till commit is successful

Question

I use Spring 2.5 and Hibernate JPA implementation with Java and "container" managed Transactions.

I have a "after user commit" method that updates data in background and need to be committed regardless of ConcurrencyFailureException or StaleObjectStateException exception, because it will never be shown to client. In other words, need to make Optimistic Lock to Pessimistic. (Could happen if methods execution will take little bit longer and someone changed data in other transaction)

---

I read a a lot about idempotent stuff, retry if exception in search for DEFAULT_MAX_RETRIES or 6.2.7. Example or chapter 14.5. Retry. I also found in stackoverflow here and here.

I tried this:

public aspect RetryOnConcurrencyExceptionAspect {

    private static final int DEFAULT_MAX_RETRIES = 20;
    private int maxRetries = DEFAULT_MAX_RETRIES;

    Object around(): execution( * * (..) ) && @annotation(RetryOnConcurrencyException) && @annotation(Transactional) {

        int numAttempts = 0;
          RuntimeException failureException = null;
          do {
                numAttempts++;
                try {
                    return proceed(); 
                } 
                catch( OptimisticLockingFailureException ex ) {
                    failureException = ex;
                }
                catch(ConcurrencyFailureException ex) {
                    failureException = ex;
                }
                catch( StaleObjectStateException ex) {
                    failureException = ex;
                }
          } while( numAttempts <= this.maxRetries );
          throw failureException;

    }
}

RetryOnConcurrencyException is my Annotation to mark methods that need to be retried, if a exception occurrs. Didn't work... I also tried several ways like SELECT ... FOR UPDATE, EntityManager.lock(...)

What is the best way to avoid stale data, dirty reads etc. such a strategy with Spring? Retry?, synchronized?, JPA lock?, isolation?, select ... for update? I could not get it to work and I really happy about any help.

---

Here is some pseudo code what I like to do:

void doSomething(itemId) {
    select something into A;
    select anotherthing into B;

    // XXX
    item = getItemFormDB( itemId ); // takes long for one user and for other concurrent user it could take less time
    item.setA(A);
    item.setB(B);

    // YYYY
    update item; 
}

Between // XXX and // YYY another session could modify the item, then the StaleObjectStateException gets thrown.

Answer 1

I got a solution but I think it's ugly. I catch all RuntimeException and it only works for new transactions. Do you know how to make it better? Do you see any problems?

First, I made an Annotation:

@Target({ElementType.METHOD})
@Retention(RetentionPolicy.RUNTIME)
public @interface RetryingTransaction {
     int repeatCount() default 20;
}

Then I made a interceptor like this:

    public class RetryingTransactionInterceptor implements Ordered {
      private static final int DEFAULT_MAX_RETRIES = 20;
      private int maxRetries = DEFAULT_MAX_RETRIES;
      private int order = 1;

      @Resource
      private PlatformTransactionManager transactionManager;

      public void setMaxRetries(int maxRetries) {
          this.maxRetries = maxRetries;
      }
      public int getOrder() {
          return this.order;
      }
      public void setOrder(int order) {
          this.order = order;
      }

      public Object retryOperation(ProceedingJoinPoint pjp) throws Throwable {
          int numAttempts = 0;
          Exception failureException = null;
          do {
                numAttempts++;
                try {
                    DefaultTransactionDefinition def = new DefaultTransactionDefinition();
                    def.setPropagationBehavior(TransactionDefinition.PROPAGATION_REQUIRES_NEW);
                    TransactionStatus status = transactionManager.getTransaction(def);

                    Object obj = pjp.proceed();

                    transactionManager.commit(status);      

                    return obj;
                } 
                catch( RuntimeException re ) {
                    failureException = re;
                }
          } while( numAttempts <= this.maxRetries );
          throw failureException;
      }
}

Spring applicationConfig.xml:

<tx:annotation-driven transaction-manager="transactionManager" order="10" />

<bean id="transactionManager" class="org.springframework.transaction.jta.JtaTransactionManager">
    <property name="transactionSynchronizationName">
        <value>SYNCHRONIZATION_ALWAYS</value>
    </property>
</bean>

<bean id="retryingTransactionInterceptor" class="com.x.y.z.transaction.RetryingTransactionInterceptor">
    <property name="order" value="1" />
</bean>

<aop:config>
    <aop:aspect id="retryingTransactionAspect" ref="retryingTransactionInterceptor">
        <aop:pointcut 
            id="servicesWithRetryingTransactionAnnotation" 
            expression="execution( * com.x.y.z.service..*.*(..) ) and @annotation(com.x.y.z.annotation.RetryingTransaction)"/>
        <aop:around method="retryOperation" pointcut-ref="servicesWithRetryingTransactionAnnotation"/>
    </aop:aspect>
</aop:config>

And a method annotated like this:

@RetryingTransaction
public Entity doSomethingInBackground(params)...

Answer 2

Use Spring Retry to retry whole method when a version number or timestamp check failed (optimistic lock occurs).

Configuration

@Configuration
@EnableRetry
public class FooConfig {
     ...
}

Usage

@Retryable(StaleStateException.class)
@Transactional
public void doSomethingWithFoo(Long fooId){
    // read your entity again before changes!
    Foo foo = fooRepository.findOne(fooId);

    foo.setStatus(REJECTED)  // <- sample foo modification

} // commit on method end

Project configuration

Spring Boot application has defined valid spring-retry version, so only this is required:

<dependency>
    <groupId>org.springframework.retry</groupId>
    <artifactId>spring-retry</artifactId>
</dependency> 

Answer 3

We have this and what we do is:

1. Flush the session (to make sure the upcoming update will be the only one queued)
2. Load the instance
3. Do the change

4. On StaleObjectStateException, clear the action queue

   ((EventSource) session).getActionQueue().clear()
   

   and retry from #2

We have a retry counter to re-throw the exception in the end.

NOTE: This is not an officially supported method (Hibernate clearly states that a session which has thrown an exception should be discarded and not re-used), but it's a known work-around (with the limitation that you can't selectively remove the update action, but must clear the whole queue).

Answer 4

Throwing out another option here: BoneCP (http://jolbox.com) has support to automatically retry transactions upon failure (including when DB goes down, network fails, etc).