A collaborative scheduling approach for service-driven scientific workflow execution

Abstract

Scientific workflow execution often spans multiple self-managing administrative domains to obtain specific processing capabilities. Existing (global) analysis techniques tend to mandate every domain-specific application to unveil all private behaviors for scientific collaboration. In practice, it is infeasible for a domain-specific application to disclose its process details (as a private workflow fragment) for privacy or security reasons. Consequently, it is a challenging endeavor to coordinate scientific workflows and its distributed domain-specific applications. To address this problem, we propose a collaborative scheduling approach that can deal with temporal dependencies between a scientific workflow and a private workflow fragment. Under this collaborative scheduling approach, a private workflow fragment could maintain the temporal consistency with a scientific workflow in resource sharing and task enactments. Further, an evaluation is also presented to demonstrate the proposed approach for coordinating multiple scientific workflow executions in a concurrent environment.

Scheduling Performance Analysis on Parameter Sweep Applications in Cloud Environments

“…Let ζ denote the workload of the remote task and η denote the computing capacity of a VM which is defined as the amount of workload that can be finished in a unit time. If we assume that task arrival interval follows the Poisson distribution with parameters λ and μ, by using basic probability theory, we can obtain that With (3) and (4), we need a distribution model of ∑S i to calculate Pr{∑S i ≤ t -ζ/η} in (2). By extensive simulations, we notice that Gamma distribution is the best-fit distribution model to describe the Pr{∑S i }.…”

“…In traditional distributed systems, large-scale scientific computing applications are always the major concern as its ever-increasing resource requirements [1,2,3]. Consequently, plenty of high-performance computing platforms and systems were developed, such as cluster and grid [4,5], which are usually cost-expensive and dedicated for special groups of researchers.…”

A Novel Framework for Defining and Submitting Workflows to Service-Oriented Systems

“…Contrary to rules, roles almost lose sense within new scientific processes. In addition, activities are often related to both stateless and stateful services, unlike the common workflows of business processes where only web stateless services are composed [7]. We will now talk about service-oriented workflows several projects and approaches for modeling, composing, executing, and monitoring workflows on service-oriented systems are being developed.…”

UML Refinement for Mapping UML Activity Diagrams into BPEL Specifications to Compose Service-Oriented Workflows

“…They have to know fundamentals of their respective application areas. In order to deal with service-oriented computing, these users are constrained to master all the tools and languages employed in service technology [5]. Forcing them to learn new concepts in addition to all the specific tools of their exercising fields, increases the level of difficulty for them.…”