Grid Computing

1
Grid Computing Web services

• SNC Grid Research Group
• Robert Bram, PhD student
• Jana Dospisil

2
Applications Of Grid Computing

• Inadequate server or underutilized PCs
• Cluster idle PCs to share computing resources
• Grid computing can operate over the internet
• Need reliable and fast communication technologies
• Fast wireless networks
• Benefits to active research
• Data from a radio telescope looking for
  signatures of extraterrestrial intelligence
• Patient's response to chemotherapy
• Data relevant to prospective targets for drug
  discovery
• Processes that require heavy computation can be
  spread to underutilized machines in the network
• Virtual communities to share resources, services,
  solutions

3
Sharing Resources Over A Grid

• Mechanism to set up collaborative sessions
• Groups, organisations, departments
• Equipment belonging to the group shared by users
  on other computers
• Audio-video conferencing
• What else can we share?
• Software, software licenses, specialized
  equipment such as medical instruments,
  visualization tools
• Share computing power, resources

4
How We Share Over A Grid

• Location Transparency
• Dont need to know where the software or
  specialised equipment is
• Just that I can use it from my machine!
• Quality of Service Transparency
• Dont need to know if my machine is good enough
• Grid can get help from other (under-utilized)
  computers
• Single point of interface to all resources and
  virtual organisations
• Applications on grid, services on the grid
• Managers of resources
• Portability just need a browser!
• Monitoring progress
• Submitting jobs
• Collecting results
• Portlets are hot new Web application development
  technology
• Lack of standards makes it difficult to use over
  grid

5
Multimedia QoS Research

• Users of distributed multimedia systems have
  little or no control over their perceived quality
  of service
• Medical, teaching, robotics
• Ability to specify your expected level of service
  for all aspects of the grid (unified model of QoS
  metrics)
• User preferences, resource profiles and
  availability, user/application mobility
• Offers QoS monitoring, service tailoring and
  adaptation, active path management
• Protocol for communication of multimedia grid
  based tasks and QoS requirements
• Application of data mining techniques to
  construct optimal application configurations for
  distributed multimedia

6
Perceived Grid Configuration

• Grid nodes will have the following abilities
• Ability to accept and execute grid tasks
• In static environments
• In mobile environment (handheld, mobile phones
• Ability to download install required/updated
  components from service directory (distribution
  of services through grid)
• Support Multimedia Service Delivery
• Techniques, Mechanisms and Technologies to
  deliver MM with guaranteed QoS user perceived
  quality
• In mobile
• Highly variable in quality (and price)
• Small display

(Hopefully) High Speed Internet
7
Collaborative Research Opportunities

• It is possible to build a generic grid with
  specialised frameworks to take advantage of
  underutilized resources and parallel CPU
• Computationally intensive applications
  processing large amounts of data leading to
  discovery of new mix for good wine?
• Scientific calculations
• Monitoring and evaluating patients responses to
  drugs
• Image analysis and processing (mammograms)
• Resource scheduling, eg, machine changeover in
  wine industry
• Investigation into
• Virtual organisations on demand
• Enable collaborative processes
• Across business or departments
• Access to additional possibly remote resources
  and services, mobile access, mobile video phone
• Video streaming from a crash scene,
• Visual warnings to hospital
• Monitoring and intelligent reasoning over large
  amounts of collected data
• Indication of congested roads, shortest path
• Based on multiple sources of information, single
  portal interface