Task scheduling method for satisfying demand for reliability and used for distributed system

Abstract

The invention provides a task scheduling method for satisfying the demand for reliability and used for a distributed system. The method includes the steps that first, the reliability of all tasks in application when the tasks are executed in different processors is calculated, and sequenced from high to low; then, each task is copied one time in the available processor with the highest reliability, and the real-time reliability of each task and the real-time reliability of the application are obtained; thirdly, according to the relation between the real-time reliability and the demand for reliability, the real-time reliability of all the tasks is sequenced and copied on time in sequence, the real-time reliability of the tasks and the real-time reliability of the application are updated after copying of every time, and iterated updating is repeated until the real-time reliability of the application satisfies the demand for reliability; finally, the final reliability, the total redundantnumber and the overall operation time cost of the application are determined. The task scheduling method has the advantages of being low in redundancy and high in efficiency, and is high in practicability.

Description

technical field [0001] The invention relates to the field of distributed systems, in particular to a task scheduling method that satisfies application reliability requirements in the field. Background technique [0002] The rapid development of the computer industry has pushed various fields towards informatization and intelligence. With the rapid increase in the types and quantities of tasks, the early centralized systems can no longer meet the current increasingly high task requirements. In recent years, under the new trend of enhanced processor performance and popularization of high-speed networks, distributed systems have been widely used for their stable operation, high efficiency, and rich functions. The sharp increase of tasks is accompanied by an increase in the number of processors, and in the actual operation process, each processor may fail, and it is difficult for both the service provider and the user to monitor the working status of each processor in real time...

AI Technical Summary

Problems solved by technology

This idea usually idealizes the reliability of unallocated tasks, that is, assuming that the sub-reliability of these tasks is high enough to reduce the sub-requirements of the current task, thereby reducing the number of copies required by the current task. This type of method The advantage of is that the redundant number can be quickly determined, but the degree of optimization for the redundant number is not ideal

Another way of thinking is to tentatively copy (search) each task, then select the task that contributes the most to the improvement of application reliability for actual copying, and repeat the above operations on all tasks until the reliability of the application meets the reliability requirements. The advantage of this type of method is that it can effectively optimize the redundant number, but it takes a huge time cost and affects the execution efficiency of the system

Images