RouteGA: A Grid Load Balancing Algorithm with Genetic Support

Abstract

Motivated by the first process allocation limitations of the original Route load balancing algorithm, this paper presents RouteGA (Route with Genetic Algorithm support) which considers historical information about parallel application executions in order to optimize the first scheduling. This information is extracted by using monitors and summarized in a knowledge base used to quantify process occupation at the launch moment. Such occupation is used to parameterize a genetic algorithm responsible for optimizing the process allocation on heterogeneous computing environments such as Grids. Results confirm RouteGA overperforms the original Route, which had previously overperformed others from literature.

Grid job scheduling using Route with Genetic Algorithm support

“…These experiments allow to confirm that RouteGA carries on outperforming all the other scheduling algorithms even considering different environments, such as clusters and grids (the first approach was just evaluated using a 128-node grid in [11]). Besides observing a lower mean response time for parallel applications, the tables also allow to notice the lower standard deviation, what implies in better accuracy and stability when scheduling.…”

“…This new approach was compared to other scheduling algorithms considering a larger environment than when the first approach was proposed [11]. The results are presented in the next section.…”

“…C j is defined in (11) where PM j is the total processing cost of the task at index j , CAP d is the capacity of the computer d where the task j is allocated, and CC j is the communication cost of the task at index j to any other task (from the same parallel application) running on other computers. CC j is defined in (13) where CPS(i j , i k ) is the communication cost between the tasks at indexes j and k of the chromosome i which run, respectively, on the computers i j and i k .…”

“…The first studies to propose a genetic algorithm to optimize the first scheduling of Route considered the multi-objective fitness function (F i ) presented in (10) [11], where: MRT i is the response time of the process with the highest computational cost of a parallel application, LV i (14) is the load variance that the proposed solution causes to the environment and η is a parameter to avoid the division by zero.…”

w-TG: A Combined Algorithm to Optimize the Runtime of the Grid-Based Workflow Within an SLA Context

“…In Mello et al (2007) • While we deal with workflow, the work in Mello et al (2007) considers a group of single jobs but with no dependency among them.…”