Scheduling multi-tenant cloud workflow tasks with resource reliability

doi:10.1007/s11432-020-3295-2

Scheduling multi-tenant cloud workflow tasks with resource reliability

2022-08-29

7

PRE

AI

DOI：10.1007/s11432-020-3295-2

Original

Original request for help

Save

Abstract

En 中文

Resource reliability is crucial in scheduling workflow instances for different tenants. Both cloud resource reliability and precedence constraints in workflows bring about great challenges for these kinds of scheduling problems. In this paper, we construct a hybrid resource reliability model which is adaptively evaluated in every time window. The objective is to optimize the QoS (quality of service) for tenants which is measured by the introduced AISE (all instance success entropy) index. A scheduling algorithmic framework is proposed for the studied workflows which consider cloud resource reliability. Deadline and budget division (BD) methods are presented to divide deadlines and budgets of instances into those of tasks. A tenant sequence method is developed to determine the order of tenants. A task allocation strategy is investigated to schedule tasks that are ready to appropriate available resources. Parameters and algorithm component candidates are statistically calibrated over a comprehensive set of random instances using the analysis of variance technique. The performance of the proposed algorithm is also evaluated in practical instances.

Keywords:

multi-tenancy

resource reliability

cloud computing

scheduling

Scheduling multi-tenant cloud workflow tasks with resource reliability

Abstract

Journal

Organization