Research issues in the theme of workflow bus

1
Research issues in the theme of workflow bus
2
Outline

• Theme
• Research issues
• Plan

3
Theme A scientific workflow bus for different
e-Science domains.

• Effective reusing existing workflow systems
• Integrating solutions to different levels of
  workflow issues
• Considering workflow at the lifecycle of
  scientific experiments

4
Research issues

• Computing model related issues
• Workflow types, execution models and correlation
• Job farming in workflow system
• Component architecture related issues
• Web services in workflow coordination
• Importing different legacy tools/environments
  into workflow
• Data related issues
• Large volume data handling
• Data integration between workflows
• Data provenance
• System prototype
• Ptolemy and JADE prototype

5
Workflow types, execution models, and correlation

• To do
• Different types of workflows
• Data token
• Execution semantics
• Event handling
• Etc.
• Different execution models
• Output

6
Job farming in workflow

• To do
• Mechanisms of job farming in workflow systems
• Results
• Changes in RTMS
• Propose solutions to different application use
  cases

7
Web services in workflow coordination
Different types Of legacy components Scripts,
Java, C, R
Deployment environment

• An engineering lifecycle for developing WS.
• Survey available tools for wrapping components as
  web services
• Integrate these tools as a coherent tools for
  application scientists

A list of Existing WS Developing tools
Engineering model Phases/steps
Web services
8
Importing different legacy tools/environments
into workflow

• Wrapping tools
• Typical cases
• R
• Different solutions to include R to workflows

• Decisions on how to integrate R into workflow
  distinguish different levels of issues
• Available solutions
• Compare them

9
Data intensive scientific workflows

• Gridles, stylx etc.
• Check how data services are used in scientific
  workflow, and compare tehm

10
Data integration between workflows

• Semantic level integration
• How to integrate them

11
Data provenance

• At workflow bus level

12
Activity plan
Computing model
Web service
R in workflow
Job farming
Data intensive
Semantic integration
Prototype the system
Data provenance
Jul.
Aug.
Sep.
Oct.
Nov.
Dec.