NGG3
GRIDS
Next Generation Grids Report 2005 NGG3
Future for European Grids: GRIDs and Service Oriented Knowledge Utilities
Vision and Research Directions 2010 and Beyond
Thierry Priol IRISA/INRIA
1
NGG Group of Experts
• Information provided by a group of independent
experts convened by the European Commission with the objective to identify potential European Research priorities for Next Generation Grid(s) 2010 and beyond
• Experts are both from Industry and Academia • These are just advices to the EC from European
high-level experts
2
Next Generation Grids Reports
2005
NGG3
GRIDS
Future for European Grids: GRIDs and Service Oriented Knowledge Utilities Vision and Research Directions 2010 and Beyond
2004
NGG2
GRIDS
Requirements and Options for European Grids Research 2005-2010 and Beyond
2003
NGG1
GRIDS
European Grid Research 2005 – 2010
http://www.cordis.lu/ist/grids
3
From NGG to SOKU
Evolution of HPCN Knowledge Technologies
…
SOA Methodologies
Next
Current Grids
Generation Grids
ServiceOriented Knowledge Utility
Autonomic Computing
Evolution of the Web
Software Technologies
…
Grid is more than metacomputing plus extensions
4
Next Generation Grid definition
A total 分布式distributed dynamically reconfigurable scalable solution for business & science applications, with not only compute power but also access to information and knowledge through a coordinated set of services
5
Service definition
• An evolution rather than revolution in software
development methodologies using modular design:
Object-oriented by inheritance继承 Component-oriented by composition Service-oriented by coordination/安排orchestration at runtime
• Make the development of distributed applications more
« agile » by:
Specifying an interface contract independent for the underlying platform (HW, OS, comm. protocol, languages) Dynamic discovery and service invocation through messages Maintaining its own state (self-contained) Loose coupling of services / 空中on-the-fly network binding Tolerating evolution at runtime Web Service is just a technology to implement services
• But service does not mean “Web Service”
6
Next Generation Grids Report 2003: Vision
“Next Generation Grid(s) 1 - European Grid Research 2005 - 2010”, June 2003
Simplification
Next Generation Grids
Abstraction
Architectural Vision
• Societal behaviour (millions of self-organising nodes) • Computational semantics, ontologies, meta-descriptions • Pervasive virtual organisations
Virtualization
7
Next Generation Grids Report 2003:
Identified Research Themes
Open Reliable Scalable Persistent Transparent Person-centric Pervasive Secure / trusted Standards-based
Properties
Research Themes
Virtual Organisation Systems Management Co-ord. and orchestration Information representation Next Generation Grid(s)
Facilities
User Interface Grid Economies Business models
Models
8
Next Generation Grids Report 2004
“Next Generation Grids 2 – Requirements and Options for
European Grids Research 2005–2010 and beyond”, August 2004
Driving Scenarios
Crisis Management – Pro-active PDA
Research Topics
Mobile and Embedded Grids
NetworkCentric (Grid) Operating Systems
Mastering Grid Complexity
Next Generation Grids
Architectural Vision
NGG1 vision and research challenges
En dUs er V
Open Reliable Scalable Persistent Transparent Person-centric Pervasive Secure / trusted Standards-based
ar ftw So
isi on
eV isio n
Research Themes
Next Generation Grid(s)
Virtual Organisation Systems Management Co-ord. and orchestration Information representation
User Interface Grid Economies Business models
9
Next Generation Grids Report 2005
Next Generation Grids 3 Future for European Grids: GRIDs and Service Oriented Knowledge Utilities – Vision and Research Directions 2010 and Beyond, December 2005
Service-Oriented Knowledge Utility (SOKU)
A flexible, powerful and cost-efficient way of building, operating and evolving IT intensive solutions for business, science and society.
building on existing industry practices, and emerging technologies support ecosystems that promote collaboration and self-organization towards increased agility, lower TCO, broader availability of services for all empowering service providers, integrators and end-consumers of ICT (r)evolution of concepts from Web, Grid & Knowledge technologies as safe, ease und ubiquitous as existing utilities such as electricity or water
10
Service-Oriented Knowledge Utility
The architecture comprises services which may be instantiated and assembled dynamically, hence the structure, behaviour and location of software is changing at run-time;
Services are knowledgeassisted („semantic‟) to facilitate automation and advanced functionality, the knowledge aspect reinforced by the emphasis on delivering high level services to the user
A utility实用程序 is a directly and immediately useable service with established functionality, performance and dependability, illustrating the emphasis on user needs and issues such as trust
11
From Web towards SOKU
Web
Methodologies
Service Oriented Architecture
Grid Stateful service Utility Agent Technologies 自动Autonomic Stateful Service Utility Semantics Societal Autonomic Stateful Service Utility Heuristics试探法 Knowledge aware Societal Autonomic Stateful Service Utility Reliable Knowledge aware Societal Autonomic Stateful Service Utility = SOKU
Formal approaches
Next Generation Grids and SOKU
12
Next Generation Grids Report 2005
Future for European Grids: GRIDs and Service Oriented Knowledge Utilities – Vision and Research Directions 2010 and Beyond, December 2006
Driving Scenarios
Business/Enterprise - End-User - Manufacturing/Industrial
Research Topics
Lifecycle Management
Trust and Security in Virtual Organizations
Adaptability, Scalability, Dependability
Raising the Level of Abstraction
Pervasiveness and Context Awareness of Services
Semantic Technologies
Human Factors and Societal Issues
Next Generation Grids
Architectural Vision
NGG1&NGG2 vision and research challenges
En dUs er V
Open Reliable Scalable Persistent Transparent Person-centric Pervasive Secure / trusted Standards-based
ar ftw So
isi on
eV isio n
Research Themes
Next Generation Grid(s)
Virtual Organisation Systems Management Co-ord. and orchestration Information representation
User Interface Grid Economies Business models
13
Lifecycle Management
• On the fly creation of services
Deployment, migration on heterogeneous resources
• Robust, efficient and semantically aware service discovery
versioning support, peer-to-peer approaches
• Composition of services based on semantics
Orchestration / Choreography
• Management of functional and non-functional properties
performance, quality of service (QoS), dependability and security aspects.
• Support multiple economy models
support reliable and scalable accounting, billing, secure access to resources
14
Adaptibility, Scalability & Dependability
• Adaptability • Scalability
Self-* systems is central to SOKU
How to handle millions of entities (services, resources, …) SOA means services and thus servers…
Using non-centralized methods (from client/server to peer-topeer)
• Dependability
Especially important for mission or life-critical services How to deliver dependable systems with acceptable performance ?
15
Raising the level of abstraction
• Web services are just an instruction set of a SOKU
computing infrastructure
• Higher level programming models and tools are
required
• Rapid development, prototyping, debugging of
generic SOKU services
16
To give a practical example…
Next Generation Grids
Non-Grid Computer
Languages, Tools, Environments
Operating Systems
High-Level Programming languages, tools and systems
snr impuv ne o im ) dckpe o TF oct pae Qoqiseruionneer sé otmese n ( r mes. é eus t I u Tr ™ is ic F t g
Qui ckTime™ et un décompresseur TIF F (non com pressé) sont requis pour visionner cet t e image.
Qu i k T me ™ et u n c i d éc om pre s se ur T I F (n on co mp res s é) F s on t re qu i po ur v i i n ne r ce t e i mag e . s so
Q u kTim ic e™ et un dé mpre ur TIF F (n c co sse on omp ress é) so req s po vi io er ce e ima . nt ui ur s nn tt ge
QuickTime™ et un décompresseur TIFF (non compressé) sont requis pour visionner cette image.
?
WSRF
Q u Time et un ick ™ dé o c mpress e TIF F (no c o r e sé) ur n mp s s o req po v i io er c ette ima nt uis ur s nn ge.
Qui ckTime™ et un décompresseur TIF F (non com pressé) sont requis pour visionner cet te image.
QuickTime™ et un décompresseur TIFF (non compressé) sont requis pour visionner cette image.
QuickTime™ et un dé co mpresseur TIFF (no n comp re ssé) sont re quis po ur visionne r ce tte imag e.
Quic k ime™ et un T déco mpr esseur T IF F ( non compr essé ) sont req uis pour v ionner cette imag e. is
QuickTime™ et un décompresseur TIFF (non compressé) sont requis pour visionner cette image.
_main: .stabd 68,0,2 stmw r30,-8(r1) stwu r1,-80(r1) mr r30,r1 .stabd 68,0,4 LBB2:
WSDL
< portType name="calcPortType" > < /portType >
Instruction Set
LBE2:
lfd f13,32(r30) lfd f0,40(r30) fadd f0,f13,f0 stfd f0,48(r30) .stabd 68,0,5 mr r3,r0 lwz r1,0(r1) lmw r30,-8(r1) blr
PowerPC IA-32 IA-64 …
Quic k ime™ et un T déco mpr esseur T IF F ( non compr essé) sont req uis pour v ionner cette imag e. is
nu te ™emi T ciuQ k )é sse rpmoc non ( F FI T r uesse rpm océd .e gami et tec rennois v ruop siu qer tnos i
Hardware Resources
Qui ckTime™ et un décompresseur TIF F (non com pressé) sont requis pour visionner cet te image.
Qu ickTime™ et un déc omp resse ur TIFF (n on comp ress é) son t req uis pou r v isionn er cette image.
Internet
17
Conclusion
• SOKU will shape the future of Grid research in
Europe
It extends the Grid vision to cover a broader range of applications and thus a wider impact on economy It gives more challenges to researchers to realize this vision
• Proposers for call6 are invited to consider the
NGG3 report
18
Acknowledgments
• Special thanks to
W. Boch and his team (European Commission Unit F2) Keith Jeffery (NGG3 chairman) David de Roure (NGG3 editor-in-chief) All NGG3 experts
19