modehuan opened a new issue #1850: Ensure that each cycle of the task runs properly URL: https://github.com/apache/incubator-dolphinscheduler/issues/1850 **Is your feature request related to a problem? Please describe.** Scenario: the task was completed and executed in T days on a certain day, but the task generated in T+1 days was not normally distributed in the system upgrade yesterday. After the normal operation of T+2 system,The system only assigned T+2 days of tasks, missing T+1 days of tasks. In the case of incremental data or daily cumulative tasks. That leads to missing T1 days of data. And all T+1 day missions that depend on that task will fail. Failed to execute successfully. **Describe the solution you'd like** The cycle of the last task execution should be checked when each task is dispatched. 1. If the last mission is successful, start the distribution from the last successful mission New cycle task. Until the latest cycle. 2. If the last task execution fails, the new task will not be distributed, and the new task will be distributed successively after the successful processing of the failed task. All tasks that depend on it go into wait mode. This ensures that all tasks are guaranteed at each ring point. **你的功能要求是否与问题有关?请描述* * 场景：某日任务在Ｔ天完成执行，但是昨天系统升级没有正常派发生成Ｔ+１日的任务，在Ｔ+２系统正常运行后， 系统只派发了Ｔ+２日的任务，缺失Ｔ+１日的任务．在增量数据或者日累计的任务的情况下．就导致缺失Ｔ１日的数据． 而且依赖该任务的所有任务的Ｔ+１日的任务都会失败．无法执行成功． **描述你想要的解决方案** 每个任务派发时应该检查最后一次任务执行的周期．１.如果最后一次任务执行成功，则依次从最后一次成功的任务开始派发 新的周期任务．直到最新周期．２.如果最后一次任务执行失败，则不派发新任务，待失败任务处理成功后再依次派发新任务， 所有依赖它的任务都进入等待状态．这样才会保证所有的任务每个环点都有保证．
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
