Structured Composition of Dataflow and Control-Flow for Reusable and Robust Scientific Workflows

PDF Version Also Available for Download.

Description

Data-centric scientific workflows are often modeled as dataflow process networks. The simplicity of the dataflow framework facilitates workflow design, analysis, and optimization. However, some workflow tasks are particularly ''control-flow intensive'', e.g., procedures to make workflows more fault-tolerant and adaptive in an unreliable, distributed computing environment. Modeling complex control-flow directly within a dataflow framework often leads to overly complicated workflows that are hard to comprehend, reuse, schedule, and maintain. In this paper, we develop a framework that allows a structured embedding of control-flow intensive subtasks within dataflow process networks. In this way, we can seamlessly handle complex control-flows without sacrificing the ... continued below

Physical Description

PDF-file: 8 pages; size: 0.5 Mbytes

Creation Information

Bowers, S; Ludaescher, B; Ngu, A & Critchlow, T September 7, 2005.

Context

This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided by UNT Libraries Government Documents Department to Digital Library, a digital repository hosted by the UNT Libraries. More information about this article can be viewed below.

Who

People and organizations associated with either the creation of this article or its content.

Publisher

Provided By

UNT Libraries Government Documents Department

Serving as both a federal and a state depository library, the UNT Libraries Government Documents Department maintains millions of items in a variety of formats. The department is a member of the FDLP Content Partnerships Program and an Affiliated Archive of the National Archives.

Contact Us

What

Descriptive information to help identify this article. Follow the links below to find similar items on the Digital Library.

Description

Data-centric scientific workflows are often modeled as dataflow process networks. The simplicity of the dataflow framework facilitates workflow design, analysis, and optimization. However, some workflow tasks are particularly ''control-flow intensive'', e.g., procedures to make workflows more fault-tolerant and adaptive in an unreliable, distributed computing environment. Modeling complex control-flow directly within a dataflow framework often leads to overly complicated workflows that are hard to comprehend, reuse, schedule, and maintain. In this paper, we develop a framework that allows a structured embedding of control-flow intensive subtasks within dataflow process networks. In this way, we can seamlessly handle complex control-flows without sacrificing the benefits of dataflow. We build upon a flexible actor-oriented modeling and design approach and extend it with (actor) frames and (workflow) templates. A frame is a placeholder for an (existing or planned) collection of components with similar function and signature. A template partially specifies the behavior of a subworkflow by leaving ''holes'' (i.e., frames) in the subworkflow definition. Taken together, these abstraction mechanisms facilitate the separation and structured re-combination of control-flow and dataflow in scientific workflow applications. We illustrate our approach with a real-world scientific workflow from the astrophysics domain. This data-intensive workflow requires remote execution and file transfer in a semi-reliable environment. For such work-flows, we propose a 3-layered architecture: The top-level, typically a dataflow process network, includes Generic Data Transfer (GDT) frames and Generic remote eXecution (GX) frames. At the second level, the user can specialize the behavior of these generic components by embedding a suitable template (here: transducer templates for control-flow intensive tasks). At the third level, frames inside the transducer template are specialized by embedding the desired implementation. Our approach yields workflows that are more robust (fault-tolerance strategies can be define by control-flow driven transducer templates) and at the same time more reuseable, since the embedding of frames and templates yields more structured and modular workflows.

Physical Description

PDF-file: 8 pages; size: 0.5 Mbytes

Source

  • Presented at: Symposium on Applied Computing, Dijon , France, Apr 23 - Apr 27, 2006

Language

Item Type

Identifier

Unique identifying numbers for this article in the Digital Library or other systems.

  • Report No.: UCRL-CONF-215235
  • Grant Number: W-7405-ENG-48
  • Office of Scientific & Technical Information Report Number: 883516
  • Archival Resource Key: ark:/67531/metadc892383

Collections

This article is part of the following collection of related materials.

Office of Scientific & Technical Information Technical Reports

Reports, articles and other documents harvested from the Office of Scientific and Technical Information.

Office of Scientific and Technical Information (OSTI) is the Department of Energy (DOE) office that collects, preserves, and disseminates DOE-sponsored research and development (R&D) results that are the outcomes of R&D projects or other funded activities at DOE labs and facilities nationwide and grantees at universities and other institutions.

What responsibilities do I have when using this article?

When

Dates and time periods associated with this article.

Creation Date

  • September 7, 2005

Added to The UNT Digital Library

  • Sept. 23, 2016, 2:42 p.m.

Description Last Updated

  • Dec. 2, 2016, 12:53 p.m.

Usage Statistics

When was this article last used?

Yesterday: 0
Past 30 days: 0
Total Uses: 2

Interact With This Article

Here are some suggestions for what to do next.

Start Reading

PDF Version Also Available for Download.

International Image Interoperability Framework

IIF Logo

We support the IIIF Presentation API

Bowers, S; Ludaescher, B; Ngu, A & Critchlow, T. Structured Composition of Dataflow and Control-Flow for Reusable and Robust Scientific Workflows, article, September 7, 2005; Livermore, California. (digital.library.unt.edu/ark:/67531/metadc892383/: accessed September 23, 2018), University of North Texas Libraries, Digital Library, digital.library.unt.edu; crediting UNT Libraries Government Documents Department.