Cross-cluster task execution method and device, computer equipment and storage medium

Abstract

The invention relates to a cross-cluster task execution method and device, computer equipment and a storage medium. The method comprises the steps of obtaining a dependency relationship between tasksin at least two clusters and the starting time of a bottommost sub-level task, and performing topological sorting on the dependency relationship between the tasks to obtain dependency relationship structure data; triggering the bottommost child-level task according to the starting time, and checking a parent-level task execution state corresponding to the bottommost child-level task according to the dependency relationship structure data; after all parent-level tasks are successfully executed, submitting child-level tasks in advance; checking an execution state of a task on which a pre-submitted sub-level task depends by utilizing a local scheduling service and a remote scheduling service; and after the local parent-level task and/or the remote parent-level task are/is successfully executed, submitting the current child-level task to be executed. By adopting the method, whether the associated task is executed or not can be accurately estimated, execution errors of cross-cluster tasks are reduced, and the resource utilization rate is increased.

Description

technical field [0001] The present application relates to the field of computer technology, in particular to a cross-cluster task execution method, device, computer equipment and storage medium. Background technique [0002] A cluster is a computer group that is interconnected through a network, a group of independent computers, and can be managed in a unified manner. The cluster has a high computing speed and can complete calculations with a large amount of computing; it has high responsiveness and can reduce the overall operation and maintenance cost. Therefore, clusters are used more and more. But task execution between clusters has always been a problem that needs to be solved. [0003] Since the dependencies between tasks cannot be directly set between multiple clusters, it is necessary to regularly check whether the associated tasks have been executed. For example, the A1 task under the A cluster cannot be executed until the B1 task under the B cluster is completed....

AI Technical Summary

Problems solved by technology

Since it is impossible to accurately estimate the execution completion time of the B1 task, if the A1 task starts to execute befo

Images