具有完成任务的服务结构任务队列 [英] Service fabric task queue with completion task

问题描述

我有一份工作可以分解为一些较小的任务，每个任务最多可能需要30分钟才能完成.任务完成后，必须运行清理任务.

I have a job that can be broken down into smaller tasks, each that may take up to 30 minutes to complete. After the tasks are complete, then a cleanup task must be run.

每个任务都是CPU密集型操作，必须在单独的计算机上运行.

Each Task is a CPU intensive operation that must be run on a separate machine.

问题是我无法找到一种方法来知道所有任务何时完成.如何在Service Fabric中创建应用程序，以便在完成所有任务后可以运行其他清理任务?

The problem is that I can not find a way to know when all tasks are complete. How can I create an application in Service Fabric, such that when all tasks are complete, an additional cleanup task can be run?

此外，如何使它在任何任务失败或意外关闭时，向所有工作程序发送取消信号，然后调用清除任务?

Additionally, how can I make it such that when any of the tasks fail or unexpectedly shut down, a cancellation signal is sent to all workers, and then call the cleanup task?

推荐答案

您也可以根据需要使用ServiceFabric Actors.

You could also use ServiceFabric Actors for your requirement.

• 创建一个触发该任务的中央"JobService"(例如，托管WebApi的无状态服务)
• 创建一个"TaskActor"来处理单个任务
  • 在ActorInterface中发布一个cancellishToken
  • Create a central "JobService" that is triggered to do the work (e.g. a stateless service that hosts a WebApi)
  • Create a "TaskActor" that will handle a single task
    • Publish a cancellationToken in the ActorInterface

    您可以看一下分区概念，以了解参与者如何在您的节点上分布.

    You could take a look at the partition concept to see how the actors are spread across your nodes.

    https://docs.microsoft.com/zh-cn/azure/service-fabric/service-fabric-reliable-actors-platform#service-fabric-partition-concepts-for-actors

    根据您的情况，actorId可以保持一致(例如taskActor1，taskActor2)以重复使用资源，也可以使用GUID使它们对于每个作业都是唯一的.

    Depending on your scenario, the actorIds could be consistent (e.g. taskActor1, taskActor2) to reuse the resources, or you could use a guid to make them unique for every job.

    这篇关于具有完成任务的服务结构任务队列的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！