International Telecommunication Union
ITU-T
What is a Grid?
Dave Berry
Deputy Director, Research & Einfrastructure Development
National e-Science Centre, UK
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
Computing as a Commodity
ITU-T
o From hand-built research
o
o
o
o
computers…
From individual computers…
From specialist
supercomputers…
From proprietary formats…
From individual servers…
o to PC’s, PDAs and mobile
o
o
o
o
o From applications…
o
o From ownership…
o
o From silos…
o
phones
to the Internet and the WWW
to clusters and cyclescavenging
to standards and ontologies
to virtualised, dynamically
provisioned server farms
to services
to computing-on-demand
to Grids
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
2
What is a Grid?
ITU-T
o A grid is a system consisting of
• Distributed but connected resources and
• Software and/or hardware that provides and manages
logically seamless access to those resources to meet
desired objectives
License
License
Web
Web
server
server
Handheld
Server
Workstation
Database
Database
Supercomputer
Cluster
Data Center
Printer
Printer
R2AD
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
3
Grids vs. Distributed Computing
ITU-T
o Existing distributed applications:
• Tend to be specialised systems
• Intended for a single purpose or user group
o Grids go further and take into account:
• Different kinds of resources
• Different kinds of interactions
• Dynamic nature
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
4
Ideas and Forms
ITU-T
o Key ideas
• Virtualised resources
• Secure access
• Dynamic provisioning
o Many forms
•
•
•
•
•
•
Cycle stealing
Linked supercomputers
Distributed data management
Commercial data centres
Utility computing
Collaboration Grids
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
5
Grids In Use: E-Science Examples
••Data
Datasharing
sharingand
andintegration
integration
−−Life
Lifesciences,
sciences,sharing
sharingstandard
standarddata-sets,
data-sets,combining
combining
collaborative
data-sets
collaborative data-sets
−−Medical
Medicalinformatics,
informatics,integrating
integratinghospital
hospitalinformation
information
systems
for
better
care
and
better
science
systems for better care and better science
−−Sciences,
Sciences,high-energy
high-energyphysics
physics
••Simulation-based
Simulation-basedscience
scienceand
andengineering
engineering
−−Earthquake
Earthquakesimulation
simulation
••Capability
Capabilitycomputing
computing
−−Life
Lifesciences,
sciences,molecular
molecularmodeling,
modeling,tomography
tomography
−−Engineering,
materials
science
Engineering, materials science
−−Sciences,
Sciences,astronomy,
astronomy,physics
physics
••High-throughput,
High-throughput,capacity
capacitycomputing
computingfor
for
−−Life
Lifesciences:
sciences:BLAST,
BLAST,CHARMM,
CHARMM,drug
drugscreening
screening
−−Engineering:
aircraft
design,
materials,
biomedical
Engineering: aircraft design, materials, biomedical
−−Sciences:
Sciences:high-energy
high-energyphysics,
physics,economic
economicmodeling
modeling
Slide from Hiro Kishimoto’s GGF17 Keynote
2006/10/19
6
climateprediction.net and GENIE
•
•
Response of Atlantic
circulation to freshwater
forcing
2K
Largest climate model ensemble
>45,000 users, >1,000,000
model years
10K
Large Hadron Collider
ITU-T
o The most powerful
instrument ever built to
investigate elementary
particle physics
o 10 Petabytes/year of data
o Simulation, reconstruction,
analysis:
• Requires computing power
equivalent to ~100,000 of
today's fastest PC
processors
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
8
Grids In Use: E-Business Examples
•• High-throughput
High-throughputcomputing
computing
−−Aircraft
Aircraftdesign
design
−−Drug
Drugdiscovery
discovery
−−Electronic
Electronicdesign
designautomation
automation
•• Financial
Financialservices
services
•• Large-scale
Large-scalecollaboration
collaboration
−−Portfolio
Portfoliomodeling
modeling
−−Data
integration
Data integration
−−Aircraft
Aircraftdesign
design
−−Automobile
Automobiledesign
design
•• Enterprise
EnterpriseInformation
Information
Integration
Integration(EII)
(EII)
−−Banking
Banking
−−Drug
Drugdiscovery
discovery
−−Collaborative
Collaborativeengineering
engineering
Slide from Hiro Kishimoto’s GGF17 Keynote
2006/10/19
9
Leading Grid Application Domains
ITU-T
Grid Market Opportunity 2005
Manufacturing
Financial
Services
Energy
Derivatives
Analysis
Seismic
Analysis
Statistical
Analysis
Reservoir
Analysis
Portfolio
Risk
Analysis
Mechanical/
Electronic
Design
LS /
Bioinformatics
Other
Process
Simulation
Cancer
Research
Entertainment
Finite
Element
Analysis
Drug
Discovery
Digital
Rendering
Weather
Analysis
Protein
Folding
Massive
Multi-Player
Games
Code Breaking/
Simulation
Failure
Analysis
Protein
Sequencing
Streaming
Media
Web
Applications
Academic
“Gridified” Infrastructure
Sources: IDC
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
10
Grids straddle disciplines
ITU-T
o A company working on the placement of a new factory
needs financial forecasting combined with mining of
proprietary historical data
o An industrial consortium work on the feasibility study
for a new airliner
o A crisis management team reacts to a chemical spill
using soil and weather models, demographic
information, and productivity tools for emergency
teams
Sources: Anatomy of the Grid, Foster et al.
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
11
Key Middleware Requirements
ITU-T
o Secure authorization, role and access privileges
o Execute and
manage
jobs/services
Brokering
Brokering
Service
Service
CPU
CPU
Compute
Resource
Compute
Resource
Service
Service
Data
Data
Service
Service
Application
Application
Service
Service
Printer
Printer
Service
Service
Notify
Virtualized
resources
Registry
Registry
Service
Service
Advertise
oResource
discovery,
monitoring &
control
Grid
middleware
services
Job-Submit
Job-Submit
Service
Service
o Data access,
transfer and
management
oInformation
about applications
& resources
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
12
Collaboration Grids
ITU-T
o Key concept:
• The ability to negotiate resource-sharing
arrangements among a set of participating parties
and then to use the resulting resource pool for
some purpose. (Ian Foster)
o Virtual organisations:
• Combining people and resources from different
organisations to address a given problem
• Enabled by Grid technology
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
13
Defining a Collaboration Grid
ITU-T
Three aspects:
1. Coordinating on-demand, secure access to
distributed and heterogeneous resources (cpu,
storage, bandwidth …)
2. Using standard, open, general-purpose protocols
and interfaces
3. To deliver non-trivial qualities of service
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
14
SOA & Web Services
SOA
SOA
••Flexible
Flexible
−−Locate
Locateservices
serviceson
onany
anyserver
server
−−Relocate
Relocateas
asnecessary
necessary
−−Prospective
clients
Prospective clientsfind
findservices
servicesusing
using
registries
registries
••Scalable
Scalable
−−Add
Add&&remove
removeservices
servicesas
asdemand
demandvaries
varies
••Replaceable
Replaceable
−−Update
Updateimplementations
implementationswithout
withoutdisruption
disruption
totousers
users
••Fault-tolerant
Fault-tolerant
−−On
Onfailure,
failure,clients
clientsquery
queryregistry
registryfor
foralternate
alternate
services
services
2006/10/19
Web
WebServices
Services
•• Interoperable
Interoperable
−−Growing
Growingnumber
numberofofindustry
industrystandards
standards
•• Strong
Strongindustry
industrysupport
support
•• Reduce
Reducetime-to-value
time-to-value
−−Harness
Harnessrobust
robustdevelopment
developmenttools
toolsfor
for
Web
Webservices
services
−−Decrease
Decreaselearning
learning&&implementation
implementationtime
time
•• Embrace
Embraceand
andextend
extend
−−Leverage
Leverageeffort
effortinindeveloping
developingand
anddriving
driving
consensus
on
standards
consensus on standards
−−Focus
Focuslimited
limitedresources
resourceson
onaugmenting
augmenting&&
adding
standards
as
needed
adding standards as needed
15
SOA & SOI
ITU-T
Responsiveness toto
Market
Responsiveness
Market
Business Strategy
Automation & Execution
Service
Oriented
Architecture
Business Monitoring & Analytics
Business Process
& Application Automation
Information and Data Services
Integration, Event
& Deployment Services
Collaboration & Communication
Services
Software As A Service
End-to-End,
Dynamic
Management
Access & Interface Services
IT Operations
Automation & Management
Service Level Management
& Automation
Metering, Measurement
& Charge-back
Service
Oriented
Infrastructure
Security
Infrastructure Virtualization
Operational
Operational Efficiency
Efficiency
Infrastructure
As A
Infrastructure Provisioning
ServicePlatform Management
& Monitoring
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
16
Example: Procure to Pay Process
ITU-T
Customer
Division
Shared
Service
Supplier
Outsourced
Change: Customer Order Entry
Change: Serviced Marketing, Billing, Receivables
Change: Supplier Handles Inventory
Change: Shipping by External Company
Change: Collections Outsourced
Change: Process Optimization
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
17
Motivations for Grids
ITU-T
o Scale up computing and/or data sets
o Reduce costs via capex/opex efficiencies
o Reduce time-to-results
o Provide reliability, availability
o Support heterogeneous systems & realities
o Enable collaborations
o Support a market in software services
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
18
Questions?
ITU-T
o Credits:
• Hiro Kishimoto, Fujitsu & OGF
• Dave Snelling, Fujitsu & OGF
• Franco Travostino, Nortel & OGF
• Angus McCann, IBM
• Ian Osborne, Grid Computing Now!
• Ian Foster, ANL/Globus
• IDC
ITU-T/OGF Workshop on Next Generation Networks and Grids
Geneva, 23-24 October 2006
19