Economy Grid: A New e-Paradigm for Grid/Internet Computing Special Thanks: David Abramson Jack Dongarra Wolfgang Gentzsch Jonathan Giddy Domenico Laforenza Rajkumar Buyya (buyya.com) School of Computer Science and Software Engineering Monash University, Melbourne, Australia

Presentation Online! (Updated slides)

Agenda Computing Platforms: Breaking Barriers Computing Platforms: Breaking Barriers Towards Global (Grid) Computing Towards Global (Grid) Computing – How the Grid is Different ? – Is it Internet/Web ? Next Gen. Internet ? Grid Applications ? Grid Applications ? Grid Resource Management Issues Grid Resource Management Issues Major Grid projects and Globus Major Grid projects and Globus Grid Architecture for Computational Economy (GRACE) Grid Architecture for Computational Economy (GRACE) Economic Models for Resource Trading Economic Models for Resource Trading Nimrod/G Grid Resource Broker Nimrod/G Grid Resource Broker Analogy to Electric Power Grid Analogy to Electric Power Grid Conclusions Conclusions

Computing Power (HPC) Drivers Solving grand challenge applications using computer modeling, simulation and analysis Life Sciences CAD/CAM Aerospace Military Applications Digital Biology Military Applications E-commerce/anything

2100 Desktop (Single Processor?) SMPs or SuperCom puters Local Cluster Global Cluster/Grid PERFORMANCEPERFORMANCE Computing Platforms Evolution Breaking Administrative Barriers Inter Planet Cluster/Grid ?? Individual Group Department Campus State National Globe Inter Planet Universe Administrative Barriers Enterprise Cluster/Grid ?

Killer Cluster Cluster Applications Numerous Scientific & Engineering Apps. Numerous Scientific & Engineering Apps. Parametric Simulations Parametric Simulations Business Applications Business Applications – E-commerce Applications (Amazon.com, eBay.com ….) – Database Applications (Oracle on cluster) – Decision Support Systems Internet Applications Internet Applications – Web serving – Infowares (yahoo.com, AOL.com) – ASPs (application service providers) – eChat, ePhone, eBook, eCommerce, eBank, eSociety, eAnything! – Computing Portals Mission Critical Applications Mission Critical Applications – command control systems, banks, nuclear reactor control, star-war, and handling life threatening situations.

Science Portals PAPIA PC Cluster Pentiums Myrinet NetBSD/Linuux PM Score-D MPC++ RWCP Japan:

Adoption of the Approach

Clusters of Clusters (HyperClusters) Scheduler Master Daemon Execution Daemon Submit Graphical Control Clients Cluster 2 Scheduler Master Daemon Execution Daemon Submit Graphical Control Clients Cluster 3 Scheduler Master Daemon Execution Daemon Submit Graphical Control Clients Cluster 1 LAN/WAN

Towards Grid Computing…. For illustration, placed resources arbitrarily on the GUSTO test-bed!!

Global Computational Grids (unification of geographically distributed computational and instruments)

What is Grid ? An infrastructure that couples: An infrastructure that couples: – Computers (PCs, workstations, clusters, traditional supercomputers, and even laptops, notebooks, mobile computers, PDA, and so on) … – Software ? (e.g., ASPs renting expensive special purpose applications on demand) – Catalogued Data/Databases (e.g., transparent access to human genome database) – Special Instruments (e.g., radio telescope-- Searching for Life in galaxy, for pulsars) – People/collaborators (even animals who knows ?) and offers a simple, consistent, dependable, & pervasive access across (local/wide-area) networks to present them as an unified integrated resource. and offers a simple, consistent, dependable, & pervasive access across (local/wide-area) networks to present them as an unified integrated resource.

Grid: at a glance

GRID APPLICATIONS (SKIP if TIME is LIMITED)

Grid Applications-Drivers Distributed HPC (Supercomputing) Distributed HPC (Supercomputing) – computational science. high-throughput computing high-throughput computing – large scale simulation/chip design & parameter studies Remote software access / Renting Software Remote software access / Renting Software – application service provides (ASPs) Data-intensive computing Data-intensive computing – data mining, particle physics (CERN) On-demand computing On-demand computing – medical instrumentation & network-enabled solvers Collaborative Collaborative – collaborative design, data exploration, education

P. Messina et al., Caltech SF-Express distributed interactive simulation SF-Express distributed interactive simulation 100K vehicles (2002 goal) using 13 computers, 1386 nodes, 9 sites 100K vehicles (2002 goal) using 13 computers, 1386 nodes, 9 sites Globus mechanisms for Globus mechanisms for – Resource allocation – Distributed startup – I/O and configuration – Security NCSA Origin Caltech Exemplar CEWES SP Maui SP Distributed Supercomputing (SF-Express/MPICH-G, Caltech)

Ad Hoc Mobile Network Simulation Ad Hoc Mobile Network Simulation (C. Koop, Monash): Network performance under different microware frequencies and different Weather conditions -- Used Nimrod

Image-Rendering

Challenging Issues in Grid Technology Development

Building computational grids requires – New programming tools – Software that can translate the requirements of an application into requirements for computers, networks, and storage – Security mechanisms permitting resources to be accessed only by authorized users – Computers and operating systems that are more tightly integrated with high-speed networks – And ……… strong Standardization- Harmonization EFFORTS

Domain 2 Domain 1 Grid Resource Management: Challenging Issues Ack.: globus.. Authentication (once) Specify simulation (code, resources, etc.) Discover resources Negotiate authorization, acceptable use, Cost, etc. Acquire resources Schedule Jobs Initiate computation Steer computation Access remote data-sets Collaborate on results Account for usage

Grid Components Grid Fabric Grid Apps. Grid Middleware Grid Tools Networked Resources across Organisations Computers Clusters Data Sources Scientific Instruments Storage Systems Local Resource Managers Operating Systems Queuing Systems TCP/IP & UDP … Libraries & App Kernels … Distributed Resources Coupling Services Comm. SecurityInformation … QoS Process Development Environments and Tools Languages Libraries Debuggers … Web tools Resource BrokersMonitoring Applications and Portals Prob. Solving Env. Scientific … Collaboration Engineering Web enabled Apps Resource Trading

Major GRID Projects and Initiatives

mix-and-match Object-oriented Internet-WWW Problem Solving Approach Market / Computational Economy

Many GRID Projects and Initiatives PUBLIC FORUMS – – Computing Portals – – Grid Forum – – European Grid Forum – – IEEE TFCC! – – GRID& CCGRID and more. Australia – – Nimrod/G – – EcoGrid and GRACE – – DISCWorld Europe – – UNICORE – – MOL – – METODIS – – Globe – – Poznan Metacomputing – – CERN Data Grid – – MetaMPI – – DAS – – JaWS – – and many more... Public Grid Initiatives – – Distributed.net – – – – Compute Power Market Grid USA – – Globus – – Legion – – Javelin – – AppLeS – – NASA IPG – – Condor – – Harness – – NetSolve – – NCSA Workbench – – AccessGrid – – GrADS – – and many more... Japan – – Ninf – – Bricks – – and many more...

Many GRID Testbeds... GUSTO Distributed ASCI Supercomputer NASA IPG

Globus Architecture and (3rd party) Tools Applications Core Services MDS GRAM Globus Security Interface Heartbeat Monitor Nexus Gloperf Local Services LSF CondorMPI NQEEasy TCP SolarisIrixAIX UDP High-level Services and Tools DUROCglobusrunMPI Nimrod/G MPI-IOCC++ GlobusViewTestbed Status GASS Source: Globus GRACE GARA Grid Fabric Grid Apps. Grid Middleware Grid Tools QBank eCash

Local Resource Mgr Resource Brokers Application Local Resource Mgr RSL (RSL Specialization) Information Service - MDS Resource Co-allocators

Building of a brokerage system….. Foundation for the Grid Economy

Who pays for this ??

Who pays for all this ? Any Incentive for GRID resource owners ? GUSTO Distributed ASCI Supercomputer NASA IPG

Economy Grid: GRACE Gr id A rchitecture for C omputational E conomy GRACE aims help Nimrod/G overcome the current limitations. GRACE aims help Nimrod/G overcome the current limitations. GRACE middleware offer generic interfaces (APIs) that other developers of grid tools can use along with Globus services. GRACE middleware offer generic interfaces (APIs) that other developers of grid tools can use along with Globus services.

Why Computational Economy in Resource Management ? Observe Grid characteristics and current resource management policies Grid resources are not owned by user or single organisation. Grid resources are not owned by user or single organisation. They have their own administrative policy They have their own administrative policy Mismatch in resource demand and supply Mismatch in resource demand and supply – overall resource demand may exceed supply. Markets are an effective institution in coordinating the activities of several entities. Markets are an effective institution in coordinating the activities of several entities. Traditional System-centric (performance matrix approaches does not suit in grid environment. Traditional System-centric (performance matrix approaches does not suit in grid environment. – System-Centric --> User Centric Like in real life, economic-based approach is one of the best ways to regulate selection and scheduling on the grid as it captures user-intent. Like in real life, economic-based approach is one of the best ways to regulate selection and scheduling on the grid as it captures user-intent.

Advantages of Economic-based RM System Centric --> User Centric Policy in RM System Centric --> User Centric Policy in RM Helps in regulating demand and supply Helps in regulating demand and supply – resource access cost can fluctuate (based on demand and supply and system can adapt) Scalable Solution Scalable Solution – No need of central coordinator (during negotiation) – Resources(sellers) and also Users(buyers) can make their own decisions and try to maximize utility and profit. Uniform Treatment of all Resources Uniform Treatment of all Resources – Everything can can be traded including CPU, Mem, Net, Storage/Disk, other devices/instruments – Efficient allocation of resources

Grid Node N Grid Node 2 Computational Market Model for Grid Resource Management Grid User Application Grid Resource Broker Grid Resource/Control Domains Grid Explorer Schedule Advisor Trade Manager Job Control Agent Deployment Agent Trade Server Resource Allocation Resource Reservation R1R1 Other services Grid Information Server(s) R2R2 RmRm … Charging Alg. Accounting Grid Node1 … Trading Grid Middleware … Info ? … Jobs Health Monitor

Grid Open Trading Protocols Get Connected Call for Bid(DT) Reply to Bid (DT) Negotiate Deal(DT) Confirm Deal(DT, Y/N) …. Cancel Deal(DT) Change Deal(DT) Get Disconnected Trade ManagerTrade Server Pricing Rules DT - Deal Template - resource requirements (BM) - resource profile (BS) - price (any one can set) - status - change the above values - negotiation can continue - accept/decline - validity period API

Open Trading Finite State Machine DT Offer TS DT DNDA Offer TM > > DT - Deal Template TM - Trade Manager TM - Trade Server DA - Deal Accepted DN - Deal Not accepted

Nimrod/G Resource Broker Nimrod/G Approach to Resource Management and Scheduling

A global scheduler for managing and steering task farming (parametric simulation) applications on computational grid based on deadline and computational economy. A global scheduler for managing and steering task farming (parametric simulation) applications on computational grid based on deadline and computational economy. Key Features Key Features – A single window to manage & control experiment – Resource Discovery – Trade for Resources – Scheduling – Steering & data management It allows to study the behaviour of some of the output variables against a range of different input scenarios. It allows to study the behaviour of some of the output variables against a range of different input scenarios. What is Nimrod/G ?

Nimrod/G Grid Resource Broker Architecture Grid Middleware Nimrod/G Client Grid Information Server(s) Schedule Advisor Trading Manager Nimrod/G Engine Grid Store Grid Explorer GE GIS TM TS RM & TS Grid Dispatcher RM: Local Resource Manager, TS: Trade Server RM & TS Globus,Legion, Condor-g,, Ninf,etc. G L N G Globus enabled node. Ninf enabled node. C L Condor enabled node.

A Nimrod/G Client CostDeadline Legion hosts Globus Hosts Bezek is in both Globus and Legion Domains

Change deadline/budget + Monitor activities

Active Sheets - MS Excel on the Grid! NimCache Nimrod/G

Nimrod/G Interactions Grid Info servers Resource location Queuing System Process server Resource allocation (local) User process File access I/O server Gatekeeper node Job Wrapper Computational node Dispatcher Root node Scheduler Prmtc.. Engine Trade Server

Adaptive Scheduling algorithms... LocateMachines DistributeJobs EstablishRates Meet requirements ? Deadlines and Budget Re-distributeJobs LocatemoreMachines

Resource Usage (for various deadlines)

Conclude with a comparison with the Electrical Grid……….. Where we are ????

Alessandro Volta in Paris in 1801 inside France National Institute shows the battery at the presence of Napoleon I Fresco by N. Cianfanelli (1841) (Zoological Section "La Specula" of National History Museum of Florence University)

….and in the future, I imagine a worldwide Power (Electrical) Grid …... What ?!?! This is a mad man… Oh, mon Dieu !

= 199 Years

What will be the dominant grid approach in the next future ??

Trends It is very difficult to predict the future and this is particular true in a field such as Information Technology I think there is a world market for about five computers. Thomas J. Watson Sr., IBM Founder, 1943

Trends The time is exciting but the way is hard and long…. GRID

Conclusions The Emergence of Internet as a Powerful connectivity media is bridging the gap between a number of technologies leading to what is known as Everything on IP. The Emergence of Internet as a Powerful connectivity media is bridging the gap between a number of technologies leading to what is known as Everything on IP. Cluster-based systems have become a platform of choice for mainstream computing. Cluster-based systems have become a platform of choice for mainstream computing. A number of GRID project world-wide have been presented to explore computing trend! A number of GRID project world-wide have been presented to explore computing trend! Economic based approach to resource management is the way to go in the grid environment. Economic based approach to resource management is the way to go in the grid environment. Both sequential and parallel applications run seamless on desktops, SMPs, Clusters, and the Grid without any change. Both sequential and parallel applications run seamless on desktops, SMPs, Clusters, and the Grid without any change. Grid: A Next Generation Internet ? Grid: A Next Generation Internet ?

Further Information Cluster Computing Infoware: Cluster Computing Infoware: – Grid Computing Infoware: Grid Computing Infoware: – IEEE DS Online - Grid Computing area: IEEE DS Online - Grid Computing area: – Millennium Compute Power Grid/Market Project Millennium Compute Power Grid/Market Project – Books: Books: – High – High Performance Cluster Computing, V1, V2, R.Buyya (Ed), Prentice Hall, – – The GRID, I. Foster and C. Kesselman (Eds), Morgan-Kaufmann, IEEE Task Force on Cluster Computing IEEE Task Force on Cluster Computing – GRID Forums GRID Forums – | CCGRID 2001, CCGRID 2001, GRID Meeting - GRID Meeting -

Thank You… Any ??