@terrymanu

It is a necessary feature, I just set the label to new feature.

I can make it, Could you assign to me. @terrymanu @TeslaCN

Hi @skaic
We can discuss about how to support and what will happen if timed out.

Thanks @TeslaCN . It is a necessary feature for me , I have a idea and give me some time to write, Thanks.

Core:
核心思想：

We start a process in new thread and waiting it。 When time out, call the interrupt() of the thread And cloes() of the thread socket 。
我们通过线程启动任务后，使用一个计时器进行等待，如果超时了，就调用该线程的 interrupt 方法和该线程的套接字 close 方法。

So, Ours Job can try to find a timeout By Check isInterrupted ( Thread.currentThread().isInterrupted() ) 、 catch InterruptedException | IOException , and to close self .
因此，我们的任务就可以通过检查 线程的 isInterrupted 标记(Thread.currentThread().isInterrupted()) 、 捕获 InterruptedException | IOException 来发现超时，让任务自己可以顺利的终止执行。

代码如下：

 for (int each : items) {
            ... 
            Future<Boolean> future = executorService.submit(() -> {
                try {
                    process(jobConfig, shardingContexts, each, jobExecutionEvent);
                } finally {
                    latch.countDown();
                }
                return true;
            });
            futures.add(future);
        }

        for (Future<Boolean> future : futures) {
            try {
                future.get(10, TimeUnit.SECONDS);
            } catch (final InterruptedException ex) {
                Thread.currentThread().interrupt();
            } catch (final ExecutionException ex) {
                throw new JobSystemException(ex);
            } catch (final TimeoutException ex) {
                // call the interrupt()
                future.cancel(true);
            }
        }

Job：

public class JavaSimpleJob implements SimpleJob {

    private final FooRepository fooRepository = FooRepositoryFactory.getFooRepository();

    @Override
    public void execute(final ShardingContext shardingContext) {
        List<Foo> data = fooRepository.findTodoData(shardingContext.getShardingParameter(), 10);
        for (Foo each : data) {
            if (!Thread.currentThread().isInterrupted()) {
                fooRepository.doSomeThing(each.getId());
            }
        }
    }

}

of course , the waiting timeout code can make it better .

I don't think it is elegant that let user write the Thread.currentThread().isInterrupted() in their business code. And they can also check the timeout or interrupt by themselves just by adding a little bit of code if they had already wrote some code about Thread or concurrent.
We may need a more elegant way to do that.

Right, Thread.currentThread().isInterrupted() looks stupid.

But Java deprecated Thread.stop(), Thread.suspend() methods, and use 'Thread.Interrupt() ' send the signal to the thread.
if necessary Job need to check the signal and stop itself.

We just providing a solution to stop job-thread When it is timeout.

Job execution timeout for quartz is doing like that. see it https://www.quartz-scheduler.net/documentation/quartz-3.x/packages/quartz-plugins.html#:~:text=%23-,JobInterruptMonitorPlugin,-This%20plugin%20catches