Cost-Driven Scheduling of Grid Workflows Using Partial Critical Paths

Abstract

Abstract-Recently, utility grids have emerged as a new model of service provisioning in heterogeneous distributed systems. In this model, users negotiate with providers on their required Quality of Service and on the corresponding price to reach a Service Level Agreement. One of the most challenging problems in utility grids is workflow scheduling, i.e., the problem of satisfying users' QoS as well as minimizing the cost of workflow execution. In this paper, we propose a new QoS-based workflow scheduling algorithm based on a novel concept called Partial Critical Path. This algorithm recursively schedules the critical path ending at a recently scheduled node. The proposed algorithm tries to minimize the cost of workflow execution while meeting a user-defined deadline. The simulation results show that the performance of our algorithm is very promising.

Cost-Efficient CPU Provisioning for Scientific Workflows on Clouds

“…Existing work on task scheduling and resource provisioning addresses various optimization goals in different computing systems such as grids [1,9,20,22] and clouds [3,11,17]. Execution time and cost optimization is the focus in many studies [14,25,27].…”

Robust Scheduling of Scientific Workflows with Deadline and Budget Constraints in Clouds

“…Before presenting the algorithm, some important definitions are detailed. The critical path of a workflow is the execution path between the entry and the exit nodes of the workflow with the longest execution time [1]. Critical path determines the execution time of the workflow.…”

“…Algorithm 1 details the procedure to find the PCP of a node. Partial critical paths can be scheduled on a single resource, optimizing time and cost [1]. This algorithm decomposes the workflow into smaller groups of tasks, which helps in scheduling.…”

Distributed Workflow Scheduling Under Throughput and Budget Constraints in Grid Environments

“…Such a guarantee of service is hard to provide in a Grid environment due to its shared, heterogeneous and distributed resources owned by different organizations with their own policies and pricing mechanisms [4]. Many algorithms for deadline and budget constrained scheduling have been proposed [21][22][23][24][25][26][27]. In [25], Sakellariou et al proposed two different approaches, namely the LOSS approach and the GAIN approach to find the schedule for a given DAGstructured workflow and a given set of resources without exceeding the budget and is still optimized for overall execution time.…”

“…A deadline assignment strategy was developed to distribute the overall deadline over each task partition. Abrishami et al proposed a QoS-based workflow scheduling algorithm [23] based on the partial critical paths which first tries to map the overall critical path of the workflow such that it completes before the deadline and execution cost can be minimized, then it finds the partial critical path for each mapped task on the critical path and executes the same procedure recursively. Our work differs from the above mentioned works in several aspects: (i) we consider both throughput and budget requirements; (ii) we consider incomplete Grid environment due to network connectivity and facility accessibility; (iii) we consider resource sharing among multiple concurrent computing tasks on computing nodes or concurrent data transfers over network links.…”