从tasklet存储JobExecutionContext并在另一个tasklet中访问

Question

我有一个要求,其中一个tasklet,存储在arraylist目录中的所有文件.列表的大小存储在作业执行上下文中.稍后,在另一个步骤中从另一个tasklet访问此计数.怎么做到这一点.我试图存储在jobexecution上下文中,在运行时抛出不可修改的集合异常,

public RepeatStatus execute(StepContribution arg0, ChunkContext arg1)
throws Exception {
    StepContext stepContext = arg1.getStepContext();
    StepExecution stepExecution = stepContext.getStepExecution();
    JobExecution jobExecution = stepExecution.getJobExecution();
    ExecutionContext jobContext = jobExecution.getExecutionContext();
     jobContext.put("FILE_COUNT",150000);

还将stepexection引用存储在beforestep注释中.但仍然不可能.请让我知道,如何在两个tasklet之间共享数据.

Answer 1

你有至少4种可能性:

1. 使用ExecutionPromotionListener 将数据传递给将来的步骤
2. 使用(spring)bean来保存步骤间数据,例如ConcurrentHashMap
   • 无需进一步操作,重新启动将无法访问此数据
3. 访问您的tasklet中的JobExecutionContext,应谨慎使用,这将导致并行步骤的线程问题
4. 使用新的jobcope(春季批次3推出)

从Tasklet访问JobExecution的代码示例:

1. 设置一个值

   public class ChangingJobExecutionContextTasklet implements Tasklet {
   
       /** {@inheritDoc} */
       @Override
       public RepeatStatus execute(StepContribution contribution, ChunkContext chunkContext) throws Exception {
           // set variable in JobExecutionContext
           chunkContext
                   .getStepContext()
                   .getStepExecution()
                   .getJobExecution()
                   .getExecutionContext()
                   .put("value", "foo");
   
           // exit the step
           return RepeatStatus.FINISHED;
       }
   
   }
   

2. 提取一个值

   public class ReadingJobExecutionContextTasklet implements Tasklet {
   
       private static final Logger LOG = LoggerFactory.getLogger(ChangingJobExecutionContextTasklet.class);
   
       /** {@inheritDoc} */
       @Override
       public RepeatStatus execute(StepContribution contribution, ChunkContext chunkContext) throws Exception {
           // pull variable from JobExecutionContext
           String value = (String) chunkContext
                                       .getStepContext()
                                       .getStepExecution()
                                       .getJobExecution()
                                       .getExecutionContext()
                                       .get("value");
   
           LOG.debug("Found value in JobExecutionContext:" + value);
   
           // exit the step
           return RepeatStatus.FINISHED;
       }
   }
   

我创建代码示例,在第一解决方案3 我的弹簧分批示例中找到,见模块复杂和封装interstepcommunication

Answer 2

另一种方法是使用StepExecutionListener在步骤执行后调用的方法。您的tasklet可以实现它并共享本地属性。

public class ReadingJobExecutionContextTasklet implements Tasklet, StepExecutionListener {
    private String value;

    @Override
    public ExitStatus afterStep(StepExecution stepExecution) {
        ExecutionContext jobExecutionContext = stepExecution.getJobExecution().getExecutionContext();

        jobExecutionContext.put("key", value);
        //Return null to leave the old value unchanged.
        return null;
    }
}

因此，在此步骤中，您的bean是一个tasklet和一个像下面这样的监听器。您还应该将您的范围配置为“ step”：

    <batch:step id="myStep" next="importFileStep">
        <batch:tasklet>
            <ref bean="myTasklet"/>
            <batch:listeners>
                <batch:listener ref="myTasklet"/>
            </batch:listeners>
        </batch:tasklet>
    </batch:step>

    <bean id="myTasklet" class="ReadingJobExecutionContextTasklet" scope="step">