3rd International Workshop on
Workflow Systems in e-Science

(WSES 08)

in conjunction with

8th IEEE International Symposium on Cluster Computing and the Grid

(http://staff.science.uva.nl/~zhiming/workshop/wses/,

 http://ccgrid2008.ens-lyon.fr)

 May 2008, Lyon, France

The deadline for final camera ready version is 20/Feb/08.

The page limit is 6 and it will not be possible to purchase extra pages!

Online submission is open!!!

 (CFPPDF | CFPTXT | Program | Discussion Summary | 2007 | 2006)

( new! Special session in Int’l Journal FGCS)

 

Aims and scope

 

Grid environments enable collaborations involving large numbers of people and large scale resources, and promote the emergence of a new paradigm for scientific research: e-Science. Different layers of middleware, e.g., for managing Grid resources, computing tasks, data, and information, form the basic framework for realising an e-Science environment. By automating the management of experiment routines, a scientific workflow management system hides the underlying integration details of the e-Science resources and allows a scientist to focus on the high level domain specific aspects of the experiments. The support for scientific workflows is being recognised as a crucial feature for introducing an e-Science environment to application scientists from different domains.

The WSES workshop focuses on practical aspects of scientific workflow management systems: design, implementation, applications in all fields of computational science, interoperability among workflows and the e-Science infrastructure, e.g., knowledge framework, for workflow management. The workshop aims to provide a forum for researchers and developers in the field of e-Science to exchange the latest experience and research ideas on scientific workflow management and e-Science. Live demos of workflow systems and workflow application are welcome.

The WSES 07 and WSES 06 were successfully held in the context of ICCS in Beijing in May 29 2007 and in Reading University in May 29 2006. In both years, the workshop attracted around 30 submissions. Each paper was reviewed by at least three referees, and 17 papers, including 9 regular ones and 8 short ones, were accepted. The presentations were organized as three sessions: scientific workflows applications, system architecture and middleware, and development issues. A special session for WSES 07 will appear in the International Journal of Future Generations of Computer Systems (FGCS). Selected papers of WSES 06 have appeared in a special issue of Scientific Programming Journal.

Topics

 

Authors are invited to submit original manuscripts that demonstrate current research in all areas of scientific workflow management in e-Science. The workshop solicits novel papers on a broad range of topics, including but not limited to:

  • Workflow infrastructure and e-Science middleware
  • Workflow API and graphical user interface
  • Workflow modelling techniques
  • Workflow specification language
  • Workflow execution engine
  • Dynamic workflow control
  • Workflow verification and validation
  • Workflow system performance analysis
  • Support tools for managing workflows
  • AI techniques in workflow management, e.g., planning, runtime control and user support;
  • Security control in managing workflow
  • Real-world applications of scientific workflow
  • Different levels of interoperability among workflow systems;
  • Automatic composition of scientific workflow;
  • Knowledge infrastructure in workflow management;

Paper submission and publication

The papers are limited to 6 pages each and they must follow the IEEE 8.5"x11" two-column format guidelines described at http://www.computer.org/portal/pages/cscps/cps/cps_forms.html. Workshop papers will be included in CCGrid proceedings published by IEEE Computer Society. The papers will also be available electronically in the IEEE digital library. The paper has to be submitted via online submission system. Selected best papers, after extension, will be published in a suitable international journal as a special issue.

Important Dates

  • November 12, December 16, 2007 Full paper due
  • January 3, January 20, 2008 Notification
  • January 30, February 20, 2008 Camera-ready paper due

 

Workshop

The third Int’l workshop on Workflow Systems in e-Science (WSES 2008) was successfully held in the context of Int’l Symposium on Cluster Computing and the Grid (CCGrid 08) in Lyon in May 20 2008. WSES aimed to provide a forum for researchers and developers in the field of e-Science to exchange the latest experience and research ideas on scientific workflow management and e-Science. WSES focused on practical aspects of scientific workflow management systems: design, implementation, applications in all fields of computational science, interoperability among workflows and the e-Science infrastructure, e.g., knowledge framework, for workflow management. 

 

This year, the workshop attracted 25 submissions include 5 invited papers. Each paper was reviewed by at least three referees, and 13 papers were accepted for regular talk. The presentations were organized as three sessions: data and resource management in workflow systems, execution and interactivity in workflow systems, and workflow systems. A discussion session was organized at the end of the workshop. One speaker was absent and more than 25 researchers attended the workshop.

Participants (>25):

Brice Arnauld, waleed Almodra, Khalid Belhajjame, Markus Held,  Tamas Kiss, Marco Guazzone, Bartosz Baliz, Mattoso Marta, Nguyen Toan, Marian bubak, Piote Nowakowwski, Brad ettlemyer Meltz, Daniel, Rehman Muhammed Abduk, Adam Barker, Ewa Deelman, Silvia Olabariagga, Adianto Wibisono, Tristan glatard, Johan Montagnat, Kent Wenger, Pardo Martinez, Hyeong kim, Cruz seigio, Paul roe

 

 Program

 
Session 1: Data & resource management in workflow systems
 
1.           “Metadata Management in the Taverna Workflow” 
     by Khalid Belhajjame, [slides]
 
2.           “Data Management Challenges of Large-Scale, Data-Intensive Scientific workflow”
     by Ewa Deelman, [slides]
 
3.           “Tracking and Querying in the ViroLab Virtual Laboratory”  
     by Bartosz Balis, [slides not yet available]
 
4.           “Securing Grid Workflows with Trusted Computing” 
     by Po-Wah Yau [slides not yet available]
 
5.           “A Task Pipelining Framework for e-Science Workflow Management Systems” 
     by Hyeong [slides]
 
   Session 2: Execution and interactivity in workflow systems 
 
1.            “A New Approach to Development and Execution of Interactive Applications on the Grid” 
     by Piotr Nowakowski [slides not yet available]
 
2.            “Implementation of Turing machines with the Scufl data-flow language” 
     by Tristan Glatard, [slides]
 
3.            “A Framework for Interactive Parameter Sweep Applications” 
     by Adianto Wibisono, [slides]
 
4.            “A Lightweight Middleware Monitor for Distributed Scientific Workflows”
     by Fabricio Nogueira, [slides not yet available]
 
5.            “Scheduling Dynamic Workflows onto Clusters of Clusters using Postponing Strategies” 
     by Sascha Hunold, [slides not yet available]
 
   Session 3: workflow systems
 
1.           “Comparative Studies Made Simple in GPFlow Lawrence Buckingham”, 
     by Paul Roe, [slides, not yet available not yet available]
 
2.           “Architecture of the DaltOn Data Integration System for Scientific Applications” 
     by Rahman Muhammed, [slides, not yet available]
 
3.           Discussion 
 

  Summary of the focus points of the presented papers

 
  1. Metadata and their management in the Taverna workflow system. Metadata covers the life cycle of workflow from their creation, through their use and curation until their eventual removal.
  2. Two classes of metadata: metadata that describe workflow related entities, such as services, workflows and sub-workflows, and metadata that describe workflow executions.
  3. Reproducibility, provenance, and sharing are still the critical issues for e-scientists using workflow systems. A new approach based on an Asynchronous Data Placement is proposed to improve the Workflow mapping and execution.
  4. Provenance tracking and querying to construct complex queries over provenance records. The use of ontologies for modeling provenance enables query construction in an end-user oriented manner.
  5. The use of reputation and provenance information has been proposed to avoid selecting ‘untrusted’ nodes when provisioning Grid jobs. However, this information may be unreliable or open to manipulation.
  6. The use of a Resource Broker Verification Service (RBVS) to select a trusted WRB (workflow resource broker). The approach makes use of integrity measurement, sealing and platform attestation to provide security services
  7. Tool to help the workflow systems to overlap the execution of adjacent tasks by enabling the pipelining of the intermediate data transfer between the interconnected tasks.
  8. Approach to development of interactive Grid applications used in the Application Development Platform (Appea). The environment enables the development of interactive applications, integrating  various types of Web and Grid services as well as data sources spread over a distributed infrastructure.
  9. Scufl is a turing machine complete with some restrictions. Several non trivial Scufl patterns such as self-looping or sub-workflows are required to implement Turing machine in Scufl.
  10. A framework for interactive parameter Sweep applications, the current prototype. A tuple space like workspace is used to maintain state of experiment allows basic interactivity at run time.
  11. A lightweight middleware monitoring system to design and control the parallel execution of tasks from a distributed scientific workflow. The monitoring system can be connected to a third party  workflow management system.
  12. framework responsible for implementing the data perspective of POPM (perspective oriented process modeling). The system is utilized as a plug-in for WfMS to take care about data related tasks, thus it hides the data complexity by allowing end-users to focus on their domain related tasks.
  13. Extensions to the GPFlow workflow system which facilitate interactive experimentation, automatic lifting of computations from single-case to collection-oriented computation and automatic correlation and synthesis of collections.
 

Summary of the discussion:

The discussion was very short but very focused indeed, it was clear that most of the participants had clear interest in “Data management”. When the data manipulated by a workflow become really big, the existing workflow management systems have the tendency to misbehave, and crash. One reported experience to deal with large Data management in real application showed that still a lot of manual 

tasks are involved to solve critical situations. A number of suggestions related to some concepts  presented in this workshop might be considered:

  • data cleaning presented in the talk of Ewa Deelman
  • In the Taverna approach the engine is responsible for managing and transferring "references" to data objects instead of the data objects which are often large in size.
  • easy access to distributed data vbrowser presented in the Ibis tutorial
 

 Author noticeable points

 
-              Application domains addressed in the presented papers: Bio-informatics (3 papers), medical (1 paper),
-              Any other suggestions are more than welcome.

 

Programme committee

  • Pieter Adriaans (University of Amsterdam, the Netherlands)
  • Ilkay Altintas (University of California, USA)
  • Roger Barga (Microsoft Research, USA)
  • Marian Bubak (AGH University of Science and Technology, Krakow, Poland)
  • Rajkumar Buyya (The University of Melbourne, Australia).
  • Ewa Deelman (University of Southern California, USA)
  • Thomas Fahringer (University of Innsbruck, Austria).
  • Carole Goble (University of Manchester, UK)
  • Bob Hertzberger(University of Amsterdam, the Netherlands)
  • Andreas Hoheisel (Fraunhofer Institute for Computer Architecture and Software Technology, Germany)
  • Lican Huang (Zhejiang University, China)
  • Peter Kacsuk (Hungarian Academy of Sciences, Hungarian)
  • Cees de Laat (University of Amsterdam, the Netherlands)
  • Minglu Li (Shanghai Jiaotong University, China)
  • Ling Liu (Georgia Institute of Technology, USA)
  • Shiyong Lu (Wayne State University, USA)
  • Syed Naqvi (CETIC, Belgium)
  • Maryam Purvis (University of Otago, NZ)
  • Peter Rice (European Bioinformatics Institute, UK)
  • Ian Taylor (Cardiff University, UK)
  • Zhiwei Xu (Chinese Academy of Sciences, China)

Organisers

Dr. Zhiming Zhao
email: zhiming@science.uva.nl

Tel: +31 20 5257599

Fax: +31 20 5257490

www: staff.science.uva.nl/~zhiming

Informatics Institute, University of Amsterdam
1098SJ, Amsterdam, the Netherlands

 

Dr. Adam Belloum
email: adam@science.uva.nl

www: staff.science.uva.nl/~adam

Informatics Institute, University of Amsterdam
1098SJ, Amsterdam, the Netherlands