AN ANALYSIS OF GRID TECHNOLOGIES - PowerPoint PPT Presentation

1 / 24
About This Presentation
Title:

AN ANALYSIS OF GRID TECHNOLOGIES

Description:

According to Intelligent Enterprise News (Oct. 4, 2001), 'most Web servers only ... University: Carnegie Mellon, Univ. of Chicago, Dartmouth, Univ. of Houston, USC, ... – PowerPoint PPT presentation

Number of Views:57
Avg rating:3.0/5.0
Slides: 25
Provided by: csoc
Category:

less

Transcript and Presenter's Notes

Title: AN ANALYSIS OF GRID TECHNOLOGIES


1
AN ANALYSIS OF GRID TECHNOLOGIES FOR SUPPORT OF
SPACE BASED OPERATIONS Sam Chism Shirley
Tseng LOCKHEED MARTIN SPACE OPERATIONS
INFINITE GLOBAL INFRASTRUCTURES Space Ops
2002 Houston, October 2002
2
AGENDA
The purpose of this presentation is to give a
high level introduction to grid computing,
describe its current status and identify
potential applications to NASAs space/ground
operations.
  • INTRODUCTION
  • Grid Technology Definition and Observations
  • Network Architecture and Grid Technology
  • Global Grid Forum (GGF)
  • GRIDs AND APPLICATIONS
  • Selected Current Grid Systems
  • The Information Power Grid (IPG)
  • Selected Grid Applications/Commercial Successes
  • Other Trends and Evolving Applications
  • GRID TECHNOLOGIES AND SPACE OPERATIONS
  • Grid Applications in a Generic Space Operations
    Architecture
  • SUMMARY AND CONCLUSIONS
  • GRID REFERENCES AND SOURCES (Available with
    hardcopy)

3
INTRODUCTION
  • Grid Technology Definition and Observations
  • Grid Computing is an advanced form of distributed
    computing which allows the sharing of data,
    applications, processing power and other
    resources over intranets or the public Internet.
  • These other resources include people and
    instruments, all of which can be geographically
    and/or organizationally dispersed.
  • The overall motivation for Grids is to
    facilitate the routine interactions of all of
    these resources in order to support widely
    distributed, multi-institutional science,
    engineering and/or other enterprises.
  • The central problem of grid systems is that of
    coordinating resource sharing in a distributed,
    multi-organizational environment.
  • The size of a grid can range from a relatively
    small number of local, identical computers - to a
    large, worldwide set of heterogeneous resources
  • According to Intelligent Enterprise News (Oct. 4,
    2001), most Web servers only use about 10
    percent of their processing power, while PCs use
    about 2 percent.
  • With a grid architecture, both data and program
    logic are placed at the location that optimizes
    the performance of the application as a whole.
    (Ref. 1)

4
INTRODUCTION (Contd)
iVDGL International Virtual Data Grid Laboratory
Tier0/1 facility
Tier2 facility
Tier3 facility
10 Gbps link
2.5 Gbps link
622 Mbps link
Other link
U.S. PIs Avery, Foster, Gardner, Newman, Szalay
www.ivdgl.org
Source Computing and the Globus ToolkitTM
HTML, PDF (3.2Mb), PowerPoint (3.2Mb)
5
INTRODUCTION (Contd)
Network Architecture and Grid Technology (Contd)
Operations Personnel
Development
Training Personnel
Engineering /Test
Application Portals Discipline specific tools,
etc.
Scientists /PIs
Virtual Org (VO) A
Virtual Org (VO) B
Web Services Job submission, etc.
Grid Services Grid management, etc.
Resources (Geographically Distributed)
Computer Facilities
Storage Facilities
Network Caches
Facility PCs, Servers
6
INTRODUCTION (Contd)
  • Global Grid Forum (GGF)
  • The GGF is an international community-initiated
    forum of individuals working on distributed or
    grid technologies. Its goals are
  • Facilitate/support creation of regional and
    global grids
  • Address architecture, infrastructure, standards
    and other technical requirements
  • Educate the scientific community, industry,
    government and public
  • Facilitate the application of grid technologies
  • Provide a forum for grid technologies,
    applications and opportunities
  • Exercise powers conferred under its incorporation
    (July 2001)
  • The GGFs wide membership includes
  • Government US (DOE, NASA, NSF, ..), Canada,
    Chile, Europe, Japan, ...UK
  • University Carnegie Mellon, Univ. of Chicago,
    Dartmouth, Univ. of Houston, USC, UT (Austin and
    Knoxville), Notre Dame, Purdue, Virginia Tech
  • Commercial HP, Dell, EMC, IBM, SGI, Sun, Zytec
    Telecom
  • The GGF meets 3 times a year

7
INTRODUCTION (Contd)
  • Network Architecture and Grid Technology
  • Grid Services represents a middleware layer
    between the users and resources
  • Open Grid Service Architecture (OGSA) from
    Globus/GGF (Ref. 2)
  • OGSA seeks to facilitate the development of grid
    systems by aligning Grid and Web service
    technologies to produce a common set of
    applicable standards
  • WSDL (Web Services Description Language) to
    describe web components
  • UDDI (Universal Description, Discovery and
    Integration) to find web components
  • SOAP (Simple Object Access Protocol) to bind web
    components
  • OGSA seeks to standardize the services
    interactions via use of grid service factory,
    registry, discovery, and data access.
  • OGSA adds missing functionality like workflow,
    virtualization of end systems like hosting
    environments, end-to-end QoS, change management
    via interface naming
  • Grid software is open source
  • Users may generate their own custom grid code
    systems using opens source software such as
    Globus Toolkit, or purchase ones that meet their
    needs.

8
INTRODUCTION (Contd)
Network Architecture and Grid Technology (Contd)
9
INTRODUCTION (Contd)
  • Network Architecture and Grid Technology (Contd)
  • Grid Middleware layer between the users and
    resources provides the necessary grid services
  • Grid Management manages, schedules and
    optimizes use of grid resources via uniform
    resource access
  • Brokering identifies available resources to be
    assigned current and planned activities
  • Global Queuing setting up and maintaining task
    queues to optimize performance, while maintaining
    policy agreements
  • Co-Scheduling arranging and optimizing the
    running of tasks in parallel to maximize
    throughput and performance
  • Workflow Management manages the various
    activities in conjunction with queuing and
    co-scheduling
  • Data Management manages availability and
    uniform access to data by processes
  • Collaboration supports the grid functions
    required for collaborative environments
  • Security enforces grid application policies
    through access control, and supports Virtual
    Private Networks (VPNs) and Virtual
    Organizations (VOs)

10
GRIDs AND APPLICATIONS (Contd)
  • Selected Current Grid Systems
  • US Government/Science DOE Science Grid, NASA
    IPG, TeraGrid
  • US University/Commercial Condor, Legion, Cancer
    Grid, Anthrax Grid
  • Foreign Government/Science EuroGrid, UNICORE,
    UK eScience Grid
  • The Information Power Grid (IPG) is NASA's high
    performance computational grid.
  • The IPGs vision is to revolutionize how
    computing is used in NASAs science and
    engineering by providing the middleware services
    for routinely building large-scale, dynamically
    constructed, and transient, problem solving
    environments from distributed, heterogeneous
    resources.
  • The IPG, funded by the Computing, Information and
    Communications Technology (CICT) program at NASA
    Ames Research Center, is a collaborative effort
    between NASA Ames, NASA Glenn, and NASA Langley
    Research Centers, and the NSF PACI programs at
    SDSC and NCSA
  • Selected Commercial Grid Applications/Successes
  • GlaxoKlineSmith has reduced the time for
    computational analysis of new compounds from 8 to
    2 weeks
  • Nissan Motor Co. has cut their development time
    from 36-40 to only 18 months
  • J. P. Morgan Company has claimed initial
    savings of 7 million a year by employing a
    Turbolinux EnFuzion software system to increase
    their computing efficiency.

11
GRIDs AND APPLICATIONS (Contd)
  • NASAS Information Power Grid (IPG)

NGIX
CMU
Chicago
GRC
GSFC
NREN WAN
NCSA
ARC
Testbed
HQ
LaRC
Next
JPL
MSFC
Generation
SDSC
Internet
Source Dec 2001 workshop presentation
http//www.ipg.nasa.gov/
12
GRIDs AND APPLICATIONS (Contd)
  • Other Trends and Evolving Applications
  • A number of major computer manufacturers are
    pursuing grid computing products and partnerships
    with one or more of a growing list of small
    start-ups
  • IBM and Butterfly.net have teamed to deploy the
    first of its kind commercial Grid for the online
    video gaming market (Gridcomputingplanet.com),
    enabling the cost effective delivery of
    cutting-edge video games to millions
    simultaneously
  • Sun and AVAKI have a joint effort to market and
    use the Legion grid protocol set
  • Platform Computing and SciTegic announced
    Platform Life Sciences Suite, the first
    fully-automated insilico discovery solution
  • Sun Microsystems line of grid products includes
    the Sun Grid Engine family, the iPlanet Technical
    Computing Portal, the Sun Management Center and
    Development Tools
  • More recently theyve announced their N1
    Project, focused on the future of networked
    computer systems and the use of non-procedural
    languages (e.g. Java).
  • The National Academy of Sciences and Astronomical
    Survey Committee recommended in its decadal
    survey (NAS99) the establishment of a National
    Virtual Observatory (NVO) to utilize the latest
    computer and networking technologies to connect
    the archival and real-time resources of many
    earthbound and orbital astronomical observatories
    (see http//us-vo.org )

13
GRIDs AND APPLICATIONS (Contd)
NVO Conceptual Overview
Space Based Observatories Instruments
  • Easier access to more resources,
  • allowing choices to meet needs
  • and schedules
  • Greater continuous access
  • Collection of more data for
  • later analysis
  • More wavelengths
  • Expansion across time
  • Links to other data sets,
  • related and not

Virtual Ground Station/s
Archival Storage (microwave)
Ground Based Observatories Instruments
Archival Storage (visible)
  • More reliable access to resources
  • More collaboration (RT and non-RT)
  • Networked/grid resources efficiencies
  • and flexibilities

Ground Based Observatories Instruments
Archival Storage (X-Ray)
Archival Storage (other)
University Research Facilities ()
Public Access
Government Research Facilities ()
  • More timely sharing of techniques and raw data
  • Shared computing resources for greater
    efficiency and effectivity

Includes computers, staff, local storage, etc.
14
GRID TECHNOLOGIES AND SPACE OPERATIONS (Contd)
Grid Applications in a Generic Space Ops
Architecture
Data Acquisition
User Community
Virtual Ground Station (VGS)
Enhanced Collaborative Environments
Enhanced Data Sharing
Ground Link Optimization
CSA, ESA, NASDA, NASA, RSA
WAN
Servers /WSs
Mission Monitoring and Control and IP
Facilities around the World
Mission Ops Data Storage
Grid SW
Grid SW
Virtual Computer and Storage System/s
Virtualization for Distributed Operations
15
GRID TECHNOLOGIES AND SPACE OPERATIONS (Contd)
  • Grid Virtualized Ground Stations and Ground Link
    Optimization
  • NASAs vision of satellites as nodes on the
    Internet can be implemented by an approach
    defined by web services integrated with a grid
    architecture.
  • (Noted to S. Tseng by Dr. Geoffrey Fox, CSOC
    Science Working Group, at GGF4)
  • Virtual Ground Stations and a Virtual Ground
    Network (VGN)
  • Ready access to live, local and remote sources
  • Increased capability and fault tolerance
  • Supports ground link optimization, i.e. from a
    distributed set of ground receiving sites to a
    distributed set of ground sites.

16
GRID TECHNOLOGIES AND SPACE OPERATIONS (Contd)
  • Grid Virtualization for Distributed Operations
  • Virtualized, Distributed Operations
  • Uniform, common view of mission status for more
    effective control by all parties
  • For example, the International Space Station and
    its International Partners, around the world can
    share use of Grid portals to access data (e.g.
    the IPG Launch Pad data miner (Ref. 9) and Grid
    collaborative environments for sharing of
    models, algorithms for analysis and visualization
    (Ref. 10).
  • Grid technology can be implemented so as to allow
    resources to be efficiently shared without
    compromising individual activities
  • Grid computing can facilitate the combining of
    available resources to
  • meet new objectives
  • quickly gather additional resources for
    contingency situations

17
GRID TECHNOLOGIES AND SPACE OPERATIONS (Contd)
  • Grid Computing and Enhanced Data Sharing
  • Processing of Mission/Systems, Payload Ground
    Data Augmented
  • Seamless integration with web services use of
    distributed middleware COTS and standards i.e.
    SOAP, WSDL, UDDI (Ref. 10)
  • Access to metadata service management services
  • Augment Data Distribution and Data Services
  • Remote access to distributed shared data archives
    (Ref. 10)
  • Access to metadata repositories for
    metadata-based search personalization services
  • Access to online data manipulation, analysis, and
    visualization of complex data sets (Ref. 10)
  • Use of XML based data transfer to access archives
    with different format requirements
  • Benefits Summary
  • More efficient use of available computing power
    and storage
  • Greater fault tolerance
  • More flexibility to prioritize available
    computing power, meet new requirements
  • More extensibility to meet new requirements
  • Cost savings from all of the above and from the
    reduced maintenance and sustaining.

18
GRID TECHNOLOGIES AND SPACE OPERATIONS (Contd)
  • Grid Enhanced Testing and Collaborative
    Environments
  • Test Environment (e.g. the European SpaceGrid,
    Ref. 10, slide14)
  • Remote monitoring of test.
  • Access to test data at different facilities
  • Use of Grid data management tools to integrate
    access to test data at different test archives
  • Enhanced Collaborative Environments (making
    better use of multi-media)
  • Use of Grid collaboration tools to enhance
  • Project Meetings (e.g. status and planning)
  • Working Meetings (e.g. anomaly resolutions and
    design efforts)
  • Training Sessions via the Network

19
SUMMARY AND CONCLUSIONS
  • Grids represent a major evolutionary
    architectural step in the advancement of
    networked systems, offering improved
  • Flexibility, Expandability
  • Performance and/or Throughput (Speed and/or
    Efficiency)
  • HW and SW Cost Savings
  • Maintenance and Sustaining Savings
  • Grids can extend the use of commodity network
    layers to the application layers, enabling the
    development of distributed component models that
    can be composed of distributed applications.
  • Grids represent an emerging technology. Grid
    based products have been successfully marketed
    and implemented, and new ones are continually
    being developed.
  • A grid architecture is compatible with and can be
    overlaid on existing space based data network
    architectures
  • Grids can provide solutions to many of NASAs
    processing and networking needs
  • NASAs IPG project is ideal for test-bedding grid
    applications
  • Grids WILL BE used for science processing. The
    extent of use of Grid for mission and data
    services CAN BE defined by us!

20
AN ANALYSIS OF GRID TECHNOLOGIESFOR SUPPORT OF
SPACE BASED OPERATIONS
  • REFERENCES AND BACKUP MATERIAL

21
GRID REFERENCES AND SOURCES
  • 1. An Introduction to Grid Computing,
    Peer-to-Peer and Distributed Computing
    Roberstons
  • Stephens Investment Bankers, Feb. 2001
    http//www.metistech.com/news/pdfs/Robertson_Steph
    ens.pdf
  • 2. The Physiology of the Grid, An Open Grid
    Services Architecture for Distributed Systems
  • Integration by Ian Foster, Carl
    Kesselman, Jeffrey Nick Steven Tuecke
    www.globus.org/research/papers/ogsa.pdf
  • 3. NSF Releases Middleware Package to Spur Grid
    Use, Gridcomputing.com, May 2002.
  • 4. Grid Computing M. Mitchell Waldrop, MIT
    Technology Review, May 2002.
  • 5. Applying the Lessons of Internet Services to
    Space Systems, J. Cutler, A. Fox and K. Bhasin,
    Stanford Univ. to be published at the GSAW and
    IEEE Aerospace Conferences.
  • 6. NASAs Information Power Grid W. E.
    Johnston, Arsi Vaziri, et al, NAS Div., Ames
    Research Center http//www.ipg.nasa.gov
  • 7. The Anatomy of the Grid, Enabling Scalable
    Virtual Organizations by Ian Foster, Carl
  • Kesselman Steven Tuecke 2001.
  • Internet Infrastructure Services, Introducing
    Internet 3.0 Bear Stearns, May 2001.
  • Grid for Dummies with ES Data Mining Application
    from EO Grid Workshop http//network.ceos.org/gri
    d/hinke_grid.ppt
  • SpaceGrid presentation from EO Grid Workshop
    http//network.ceos.org/grid/marchetti_grid.ppt

22
GRID REFERENCES AND SOURCES (Contd)
  • WEBSITES
  • Global Grid Forum (GGF) http//www.gridforum.o
    rg
  • Globus http//www.globus.org
  • Condor http//www.cs.wisc.edu/condor
  • National Virtual Observatory http//www.us-vo.or
    g
  • GridComputingPlanet Website http//www.gridcompu
    tingplanet.com

23
GRIDs AND APPLICATIONS
(Ref. 4) MIT Review, May 2002
24
GRIDs AND APPLICATIONS (Contd)
Write a Comment
User Comments (0)
About PowerShow.com