Introduction of Grid Technology I Management of the GRID System

1
Introduction of Grid Technology I Management
of the GRID System
Al Vincent, Director AVincent_at_its.bldrdoc.gov
Monday, September 20, 20041010 a.m. 1040 a.m.
2
ITS and NTIA

• ITS is the Research and Engineering Laboratory of
  the National Telecommunications and Information
  Administration (NTIA)
• The National Telecommunications and Information
  Administration (NTIA), an agency of the U.S.
  Department of Commerce, is the Executive Branch's
  principal voice on domestic and international
  telecommunications and information technology
  issues.
• But I do not speak for the government, these are
  my personal viewpoints

3
My Thesis

• Grid computing is a perfect Candidate for an NGN
  service
• Grid computing expansion is limited by its lack
  of complete management and control facilities
• The addition of both facilities are critical for
  its more widespread use.

4
Grid Computing is

• a type of parallel and distributed system that
  enables the sharing, selection, aggregation of
  geographically distributed resources depending on
  their availability, capability, cost, and user
  QoS requirements
• for solving large-scale problems/applications.

Rajkumar Buyya
Grid Computing and Distributed Systems (GRIDS)
Lab. Dept. of Computer Science and Software
EngineeringThe University of Melbourne,
Melbourne, Australia
5
Sample Grid Applications

• Pure Research
• Data/Decision Analysis
• Pharmaceutical Research
• Games mpMU
• Distributed Computing with Rendezvous
• Simulation
• Modeling (weather)

6
The Grid Paradigm
7
Grid Architecture (Globus)
8
Similarly

• Globus
• Gridbus
• Grace
• Nimrod-G
• Netsolve
• Albatross
• Javalin

• But these are kits and frameworks that make the
  user too aware of the connectivity, transport and
  environment and not enough about the application

9
The Challenges
10
Grid Architecture (GRACE)
Data Catalogue
Grid Bank
Information Service
Grid Market Services
Sign-on
Info ?
Grid Node N

Grid Explorer

Secure
ProgrammingEnvironments
Job Control Agent
Grid Node1
Applications
Schedule Advisor
QoS
Pricing Algorithms
Trade Server
Trading
Trade Manager
Accounting
Resource Reservation
Misc. services

Deployment Agent
JobExec
Resource Allocation
Storage
Grid Resource Broker

R1
R2
Rm
Grid Middleware Services
Grid Service Providers
11
A Grid can easily become
12
TeraGrid NCSA
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
Back on Track
18
Critical Viability Issues

• Toolkit viewpoint
• Lack of separate management of the platform of
  services
• Too much complexity on the user (managing both
  transport and service end-to-end)
• Independent Service/Resource Models
• Lack of interoperability
• Lack of standards
• Difficult to add new services openly.

19
Critical Viability Issues

• Practitioners are spending too much time as IT
  experts and too little running problems on the
  grid

20
NGN Layer(s)

• eMail
• Web Surfing
• VoIP
• Television
• GRID computing
• Data Access
• VPNs
• Distributed Business

• Optical
• Wifi / Wimax
• Cable
• xDSL
• LAN
• Beams
• RF

21
The New IT Way
22
NGN Layer(s)
23
Services and Transport

Management Plane

Control Plane

User Plane

NGN Service Stratum

Management Plane

Control Plane

User Plane

NGN Transport Stratum

24
Control and Management
25
NGN

• Services
• Resource definition
• Secure access
• Complex service definition
• Accounting
• Monitoring
• Connection of new facilities

• Technology
• Any bandwidth
• Any technologies
• New technologies added immediately
• Access from PDA/workstation

26
Moving From Distributed Computing to NGN Service

• Define external services in an open way.
• Integrate those services to the transport in a
  standard way.
• Remove external service dependence on transport
  except for simple application independant
  paradigms
• Full control and management of connections,
  service availability, QoS, etc.
• Stack everything over IPv6

27
NGN/Grid Goals

• Make it usable by sophisticated application users
  without toolkit or communications experience.
• Abstract toolkits to allow them to interoperate.
• Allow open services such as statistical
  analysis, to be brokered and sold on the net as
  black boxes.
• Allow all to become hosted managed services.

28
Conclusion

• The Grid community was going in the right IP
  direction, however modern NGN telecom
  standardized on many more things.
• If you need Grid services, the NGN model allows
  you to find and use them more easily.
• If you offer these services this model allows
  them to be used, managed and monitored as a
  service independent of toolkits.
• Thus a college or national lab can broker its
  services on the grid to multiple users on many
  differing computing basis

29
Thank you.

• Al Vincent, Director ITS
• AVincent_at_its.bldrdoc.gov
• 1 303.497.3500