Grid

1
Grid Utility Computing Do they really mean
Pervasive Services

• ICPS 2006 Panel
• Manish Parashar
• TASSL, Rutgers University,
• Piscataway, NJ USA

2
Definitely not!

• Separate communities, conferences, journals, etc.
• GGF, EGA, Grid, SC, CCGrid, HPDC, ICPS,
  Pervasive, Percom, IJGC, Ubicomp,
• Separate history/perspectives
• HPC, big science, sensor networks, real-time
• Separate demigods
• John McCarthy, Fenando Corbato, Ian Foster, Mark
  Weiser
• Separate standards bodies and standards
• IETF, OASIS, DMTF, W3C, SNIA, WSRF, OGSA, 802.XX,
  OMA,
• Separate software
• TeraGrid, Globus, Unicore, eBiquity, Gaia,
  Datafountain, Oxygen, SmartSpace, Aura,
  AccessGrid, EGEE, OSG,
• Separate sponsors, patrons, funding sources

3
So what are they about?

• Grid/Utility Computing
• Seamless, secure, on-demand access to (and
  aggregation of) geographically distributed
  computing, communication, data resources
• I.e., provide computing resources (computers,
  communication, data, instruments) as pervasive
  services
• Pervasive/Ubiquitous Computing
• Integrate computation/communication into the
  environment
• I.e., provide information and context as
  pervasive services

4
Hmm, they do have similar underlying
isssues/challenges

• Large scale distributed computing
• Service description, discovery, and composition
• Security, authentication and trust
• Data/Information management
• Dynamic assembly of components
• Dynamic adaptation to state and context
• Availability, reliability, trust,
• Management and self-management
• Standards and interoperability

5
Also,

• Grid/Utility applications have pervasive aspects
• Grid/Utility computing uses pervasive computing
  to couple with the real world
• E.g., provide dynamic (real-time) data,
  information to scientific, engineering, business
  processes
• Pervasive application have Grid/Utility aspects
• Pervasive computing uses Grid/Utility computing
  to provides services
• E.g., assimilate/process large amounts of
  information

6
The Big Picture
7
Different aspects of an overall solution?

• Applications that require
• Knowledge-based
• Information-driven
• Context-aware
• Computationally-intensive
• Distributed
• solutions!

Grid/Utility
Pervasive/ Ubiquitous
PGE
Autonomics
Semantics
Crisis management, monitor and predict natural
phenomenon, monitor and manage engineering
systems, optimize business processes
8
The key barrier

• our ability to reason are limited by our
  experiences L. Wittgenstein
• The biggest challenges are non-technical
• People from disparate fields with different
  perspectives, experiences, languages, etc., must
  work and evolve together!