Multi-copy task fault tolerance scheduling method of heterogeneous distributed system

Abstract

The invention belongs to the field of computers, and particularly relates to a multi-copy task fault tolerance scheduling method of a heterogeneous distributed system. The method includes the steps that according to the load of each task and the executing speed of each node in the system, the average executing time of all the tasks on all the processor nodes and the average communication time of all communication messages on all chains are calculated; through a bottom end priority method, the bottom end priority of any task in a task set is calculated; the tasks allowed to be scheduled are added into a scheduling queue in a priority non-increasing mode according to the priority of the tasks; the task highest in priority is selected from all the tasks allowed to be scheduled in the scheduling queue. According to the method, the execution starting time of current task scheduling copies can be further shortened, and therefore the task scheduling Makespan can be further reduced.

Description

technical field [0001] The invention belongs to the field of computers, and in particular relates to a multi-copy task fault-tolerant scheduling method in a heterogeneous distributed system. Background technique [0002] With the emergence of high-speed networks, it is feasible to connect distributed, low-cost, and most likely heterogeneous resources as a computing environment, so computers in distributed systems (such as cloud computing, grid computing, and distributed mobile computing) The heterogeneity of the system will be gradually enhanced, which provides a computing platform called a heterogeneous distributed computing (Heterogeneous Distributed Computing, HDC) system. HDC systems have become popular computing equipment for high-performance computing and information processing, and have been gradually used by key systems. HDC systems often have high system throughput and availability, and can efficiently access extensive distributed network information. HDC systems ...

AI Technical Summary

Problems solved by technology

Neither algorithm gives a scheduling reliability analysis

In the article "Reliable workflow scheduling without resource redundancy" published in the journal "Parallel Computing" in 2013, Zhao Laiping et al. proposed a fault-tolerant scheduling algorithm based on an active replication mechanism to minimize resource overhead. This algorithm selects the most reliable node to execute task copies , and does not take into account the copy completion time

In 2007, Antonios Litke et al. published the article "Efficient task replication and management for adaptive fault tolerance in MobileGrid environments" in the journal "Future Generation Computer Systems", which proposed a fault-tolerant scheduling mechanism in the mobile network computing environment, but it was aimed at independent tasks Scheduling, and the mechanism does not consider scheduling Makespan

In 2009, SungHo Chin et al. published the article "Genetic Algorithm based Scheduling Method for Efficiency and Reliability in Mobile Grid" at the conference "Ubiquitous Information Technologies & Applications", which is based on GA for task copy scheduling in a heterogeneous mobile grid environment to improve task reliability, but its for independent tasks

In 2005, Atakan Dogan et al. published the article "Biobjective scheduling algorithms for execution time-reliability trade-off in heterogeneous computing systems" in the journal "The Computer Journal". The bi-objective genetic algorithm (BGA) can optimize scheduling Makespan and reliability at the same time, but In the evolution process, BGA may produce invalid solutions that violate the inter-task dependencies

In 2011, Xiaofeng Wang et al. published the article "Optimizing the makespan and reliability for workflow applications with reputation and a look-ahead genetic algorithm" in the journal "FutureGeneration Computer Systems" using GA to satisfy task dependencies while using a two-stage strategy to simultaneously Optimize scheduling Makespan and reliability, but the algorithm does not use the task copy mechanism to improve reliability, so its reliability improvement is limited

Most of the current scheduling problems do not consider the selection of the start time position of the task copy on the node, so there are certain defects in the completion time optimization

Images