Grid Computing in Academia and Business

1
Grid Computing in Academia and Business

• Dr. Heinz Stockinger
• Research Lab for Computational Technologies and
  Applications
• Faculty of Computer Science, University of Vienna
• Rathausstrasse 19/9, A-1010 Vienna, Austria
  Email Heinz.Stockinger_at_univie.ac.at
• EC3, 3 May 2005, Vienna

2
Agenda

• General Grid Concepts
• Grids in Academia
• Main trends
• Business in the Grid (BIG) project
• Discussion

3
The Grid Vision (1)
Researchers perform their activities regardless
of geographical location, interact with
colleagues, share and access data
Grid Middleware provides part of the software
infrastructure
Scientific instruments and experiments provide
huge amount of data
4
The Grid Vision (2)

• A Grid is
• Special form of distributed computing
• Computing and storage resources are distributed
  over several locations (sites)
• Sites are typically connected via wide-area
  network links
• Site normally has a local-area network which
  itself has distributed computing and data storage
  resources
• Check list given by Ian Foster
• coordinate resources that are not subject to
  centralized control
• using standard, open, general-purpose protocols
  and interfaces
• deliver non-trivial qualities of service

5
More Grid Definitions

• coordinated resource sharing and problem solving
  in dynamic, multi-institutional virtual
  organizations. I.Foster
• A VO is a collection of users sharing similar
  needs and requirements in their access to
  processing, data and distributed resources and
  pursuing similar goals.
• Key concept
• ability to negotiate resource-sharing
  arrangements among a set of participating parties
  (providers and consumers) and then to use the
  resulting resource pool for some purpose
  I.Foster

6
The Grid distributed computing idea 1/2
7
The Grid distributed computing idea 2/2
8
Typical Grids

• The Grid vision can be applied best to
  applications that have the following features
• Distributed user community
• Lots of computing power is required
  (Computational Grid)
• Lots of storage capacity is required (Data Grid)
• Distributed storage locations etc.
• Grids can be applied in academia and industrial
  environments

9

• Currently, mainly in computing intensive
  sciences
• High Energy Physics, Earth Observation, Biology,
  Biomedicine
• Engineering, Multimedia
• Example Grid
• High Energy Physics (HEP) application

10
(No Transcript)
11
CERN (European Organization for Nuclear Research)

• Over a Petabyte of data per year
• Several thousand users

12
(No Transcript)
13
Grid Topology
14
Grid Storage Model
15
Grid Topology

• Grid computing has its roots in classical
  parallel and distributed computing domain
• Mainly designed for expensive and specialised
  parallel computing hardware
• Initially, special purpose interconnects and
  programming languages were used
• Common standards and programming models were
  required (standards like MPI and PVM were
  created)
• Standardisation also paved the way for the more
  general cluster computing approach
• High performance versus high throughput computing
  
• Originally, the parallel and distributed
  community dealt with CPU intensive applications
• Later, applications became more data intensive
  and several parallel I/O techniques were
  developed (http//www.cs.dartmouth.edu/pario)

16
Brief History of Grid Technology (2)

• The nature of distributed resources has a
  stronger impact on the Grid
• First emergence of Grid computing ideas in many
  of the early meta computing projects
• HTTP (Hyper Text Transfer Protocol)
• Made possible world-wide information sharing
• The Web that exploded in the early 1990ies can be
  considered as one of the direct predecessors of
  the Grid.
• Building on several of the Internet protocols and
  ideas from parallel and distributed computing,
  the first Grid ideas gained world-wide interest
  around 1998/99
• HTTP mainly allows for information sharing
• Grid allows for all kinds of resource sharing
• Computing and data (information) resources
• Many Grid projects have been created since that
  time
• Grid projects all over the world
• Note that world-wide distributed applications
  exist already much longer but the term Grid was
  created around 1998 by Ian Foster and Carl
  Kesselman

17
Service (Provider)
Service (Provider)
Service Broker
Service Broker
communication
communication
communication
communication
Service Requester
Service Requester
18
(No Transcript)
19

• Computing Element
• Gatekeeper (Globus)
• Condor-C (Condor)
• CE Monitor (EGEE)
• Local batch system (PBS, LSF, Condor)
• Workload Management
• WMS (EDG)
• Logging and bookkeeping (EDG)
• Condor-C (Condor)
• Storage Element
• File Transfer/Placement (EGEE)
• glite-I/O (AliEn)
• GridFTP (Globus)
• SRM Castor (CERN), dCache (FNAL, DESY), other
  SRMs

20
(No Transcript)
21

• Grid Computing mainly for computing and data
  intensive applications?
• Is there a market in business?
• Isn't that too restricted to certain specific
  domains?
• Grid Standardisation
• Global Grid Form develop standard protocols
• Open Grid Service Architecture (OGSA)
• Current Grid services are based on web services
• Often, the boarder is not clear
• In this way Grid technology can get more accepted

22

• The objective of the BIG project is to understand
  the possibilities and needs of doing business in
  the Grid infrastructure.

http//www.cs.univie.ac.at/big
23

• Market potential analysis and information
  dissemination
• Is the Grid ready for business?
• What needs to be done?
• Revisiting of existing (E-)business models for
  the Grid
• Development of novel business models for the Grid

24

• Apple Europe
• Muehlehner Tavolto
• Magna Steyr
• WIFI Wien
• MCNC (North Carolina)
• Bundesministerium für Verkehr, Innovation
  Technik
• EC3
• H3G (Drei Austria)
• Sun Microsystems Austria
• Oracle Austria
• Telekom Austria
• SAP Austria
• Bull Austria
• Uniqa
• Microsoft Austria

25

• Many multi-national IT companies have Grid
  activities
• Selling Grid products
• Often LAN and/or WAN solutions
• Participation in Grid projects or standards
• Service providers for
• CPU time
• Storage space
• Many SMEs do not yet have (detailed) Grid
  knowledge
• Technology is too new
• Still a high risk

26

• Business usually does not need the typical Grid
  domains for large storage/computing
• Needs more customer oriented services
• Web services seem to provide a better model
• However, there is no clear separation between
  Grid and web computing

27
Conclusion

• Grids are rather well established in certain
  (scientific) communities
• EU is investing lots of money into Grid
  activities
• Interested to see business industry activities
• Big companies are already investing into this
  technology
• SMEs currently only marginal players
• Further Information
• http//www.cs.univie.ac.at/cta