Metainformation and Workflow Management for Solving Complex Problems in Grid Environments

1
Metainformation and Workflow Management for
Solving Complex Problems in Grid Environments

• Han Yu, Xin Bai, Guoqiang Wang, Yongchang Ji, Dan
  C. Marinescu
• 04/26/2004

2
Outline

• Grid computing
• An infrastructure for a computational grid
• Process and case description
• Ontologies and knowledge base
• Conclusions

3
Grid Computing

• Features of a computational grid
• Large-scale
• Autonomy of individual nodes
• Heterogeneity and diversity
• Dynamic environment
• Open-ended
• Contention for resources

4
Outline

• Grid computing
• An infrastructure for a computational grid
• Process and case description
• Ontologies and knowledge base
• Conclusions

5
An Infrastructure for a Computational Grid

• Autonomous service providers
• Core services
• Provide transparent access to grid resources
• Persistent and reliable
• End-user services
• Provide computing services to grid users
• May be transient and unreliable

6
6
7
Outline

• Grid computing
• An infrastructure for a computational grid
• Process and case description
• Ontologies and knowledge base
• Conclusions

8
Process Description

• Describes the workflow, includes
• End-user activities
• Flow control activities
• Precedence relations among activities
• Data dependency among end-user activities
• Created
• Automatically by the planning service, or
• By the user

9
Flow Control Activities

• Begin, End
• Denote the start and the end of a computation
• Fork, Join
• Support concurrent execution of activities
• Choice, Merge
• Support conditional and iterative execution of
  activities

10
An example process description for 3D virus
reconstruction
10
11
Case Description

• Provides additional information for the computing
  task that a user wants to perform, including
• The set of initial data
• The expected results
• Resource requirements, constraints, preferences
• Given by users

12
Outline

• Grid computing
• An infrastructure for a computational grid
• Process and case description
• Ontologies and knowledge base
• Conclusions

13
Ontologies

• Knowledge
• Information related to the current status of the
  grid computing system
• An essential part of a service to act rationally
• Ontologies
• Specify the structure of knowledge
• Reveal relations among different concepts
• Support interoperability among different services

14
Knowledge Base

• Ontology structure
• Classes represent entities, e.g., data,
  activities
• Slots describe the attributes of an entity,
  e.g., the size and location of a data
• Instances
• Provide real instances of entities based on the
  ontology structure

15
15
16
Outline

• Grid computing
• An infrastructure for a computational grid
• Process and case description
• Ontologies and knowledge base
• Conclusions

17
Conclusions

• An intelligent infrastructure provides users
  transparent access to resources in a
  computational grid
• Coordination of computing tasks is based on
  process and case descriptions
• Ontologies support interoperability among
  different services in the infrastructure