Database Architecture for GEMLCA Resource Availability - Research Paper Example

DATABASE ARCHITECTURE FOR GEMLCA RESOURCE AVAILABILITY Grid Computing, Database Architecture for GEMLCA Resource Availability Grid Computing Grid computing is a concept that refers to the group of technologies which are used for seamless and scalable access of distributed resources. It is used for sharing, selection and aggregation of various geographically distributed computer resources. A grid client deficient in some kind of resource can access the resources of other grid clients. Computational resources as supercomputers, computer clusters, storage systems, data sources, instruments and people connected through a network are synched together and work like a single resource. Grid computing is very helpful in execution of large scale data intensive applications. GEMLCA GEMLCA was developed by the research team of Centre for Parallel Computing at the University of Westminster within the UK e-Science OGSA TestBed project. GEMLCA or Grid Execution Management for Legacy Code Architecture is a language independent solution used to deploy existing legacy code applications as a Grid service. Any application written in any language can be deployed as a grid service without the need of even legacy code i.e. binary codes and source code. GEMLCA is used to create complex Grid workflows between different grid clients and applications running in them. Functions of applications over grid can be accessed by using grid portals. Following features make GEMLCA easy to handle and deploy: It is language independent There is no need of legacy codes Existing applications can be deployed as grid services with minimal user intervention It is integrated with grid portal and workflows that enables the grid clients to access grid services GEMLCA Architecture Following diagram shows various components of a GEMLCA system: Components of GEMLCA Compute server It is a computing system consisting of PC clusters. Legacy applications are deployed on these servers as grid services. Compute servers are then accessed by various grid clients for required resources. Grid Host Environment Grid host environment is used to implement service oriented grid layer which is based on Open Grid Services Architecture (OGSA) specifications e.g. GT3, GT4, g-Lite or the OMII middleware. This layer is required to connect compute server to OSGA built grid. The selected grid should allow the grid clients to access legacy applications available in compute server like a grid or web services. Administrators should be able to deploy new legacy applications on compute server using different job managers like Condor, Fork, PBS, etc. GEMLCA Resource GEMLCA resources are consists of legacy codes which are deployed as grid services. These resources are installed on compute server. GEMLCA Resource layer manages the interaction between Grid services and Grid Host Environment. Resource Layer is used for following purposes: Deploying legacy applications as Grid services Querying the GEMLCA resources and get the list of available legacy applications getting the list of legacy parameters with default values and allow the user to modify these submitting legacy jobs to a job manager querying the status of previously submitted legacy jobs retrieving results from legacy applications GEMLCA Resource layer have following three services: GLCAdmin It deploy a legacy application as a Grid service GLCList It searches for legacy applications available over grid GLCProcess It manages the workflow processes over Grid such as submiting a legacy application using GEMLCA and getting job status and results back GEMLCA Client GEMLCA Clients are installed on client machines which can access resources from compute server. There are two types of GEMLCA clients: Command line interface Installation and usage of command line interface is difficult and requires IT skills. Graphical interface Graphical interfaces provides user friendly environment for grid clients to access grid resources. These are easy to configure and use. Grid portals are used as graphical interface for accessing grid compute servers. GEMLCA clients are built in to portals to provide grid users a high level graphical interface which enables even nave user to use resources of computational grids. Grid portals do away with needs of installing a GEMLCA client on user's machine. Any user can access the GEMLCA resources as a web service by just using a web browser. The GEMLCA Resources layer should be installed over the middleware grid client applications such as GT3 and GT4. all legacy applications which are to be shared on the grid should also be deployed as Grid services on compute server. End users can access all the resources deployed on compute server with the help of a GEMLCA grid portal. The deployment of a new legacy code on a GEMLCA implementation exposes the legacy application as a grid service to all grid clients. It looks like the legacy applications are running in its native environment in compute server. GEMLCA Resource layer handles the legacy application as Grid services and manages the communication between grid clients. The knowledge of parameters required for running legacy code and its working environment is sufficient for using the grid services. A XML-based Legacy Code Interface Description (LCID) file contains the parameter set and working environment information of legacy code. This file should be located at a predefined location so that the GEMLCA resources layer can access it to run legacy code as a grid service. Following is an example of LCID file: The LCID file consists of three sections Environment, Description and Parameters. The environment section contains the legacy code name, its executable binary file name, job manager name, maximum number of jobs that a single legacy code can submit and minimum and maximum number of processors that are required to accomplish the job. The Description section contains the legacy application in simple text format and the Parameter section contains the list of Parameters, friendly name, input/output status, order status (compulsory or optional), file or command line, and the regular expression to be used as input validation. Grid Portals Grid portals are graphical interface which visually presents all grid information like availability of resources, properties of resources, status of jobs, result of jobs etc. Portals also provide the facility of job submission. Present day portals are used to develop and submit jobs either as components of parameter studies or workflows. The use of workflow method is most common as it supports advanced form of job submission. Workflow-oriented Grid portals Workflow oriented grid portals are classified in to three categories. Following tables shows the different classes of grid portals and client portals. Multiple isolated users (MIxx) Multiple collaborative users (MCxx) Single isolated Grid (xxSI) MISI portals MCSI portals Multiple isolated Grids (xxMI) MIMI portals MCMI portals Multiple collaborative Grids (xxMC) MIMC portals MCMC portals The P-GRADE Portal P-GRADE Portal is a workflow-oriented Grid portal used to develop, execute and monitor workflow in a grid environment. It consists of high-level, graphical Web interfaces so that a person with minimal knowledge of grids and grid computing can post jobs and access resources of the grid. P-GRADE portal hides the low level details of portal thus save the user from the internal complexities of a grid environment. P-GRADE supports both sequential and parallel job submissions. Workflows created on one grid can be transported to other. Workflows are executed over computational grids according to user credentials and the portal analyses monitored trace data and gives the visual representation. P-GRADE Portal is based on GridSphere portal framework. It offers following services to its users: Creation of workflows from sequential (C, C++, Fortan, etc.) and parallel (MPI, PVM) programs Loading workflows into the portal directly from the P-GRADE Environment Executing job workflows on Globus technology based Grid resources Exploiting two-level parallelism: a. among jobs of a workflow; b. among processes inside a parallel job Parallel execution of workflow components inside one Grid or in different Grids Managing user certificates and proxy credentials to realize secure communication with grid participants Collecting trace data from the running jobs by the Mercury monitor. On-line visualization of the execution of workflows, communication between workflow components and visualization of process communication inside a parallel job running on a remote Grid site Fig: Basic architecture of a Grid system Following is listing of some features of a P-GRADE portal: Built-in graphical Workflow Editor Workflow manager to coordinate the execution of workflows in the Grid (including the coordination of the necessary file transfers) Certificate management Multi-Grid management Resource management Quota management On-line Workflow and parallel job monitoring Built-in MDS and LCG-2 based Information System management Local and Remote files handling Storage Element management JDL (Broker) support for resources of the LCG-2 Grid Workflow fault tolerance by job level rescuing Workflow archive service NGS P-GRADE Portal NGS P-GRADE portal is based on P-Grade portal and offer an alternative to UK's National Grid Service infrastructure. It is used for graphical development, execution, and visualization of sequential and parallel workflows on the NGS resources. Following screenshot shows available resources over the Grid, running processes and their status: Following are some features of NSG P-GRADE Portal: HTTPS-based access to the portal. Accessing MyProxy servers to manage user certificates and proxy credentials Developing workflows from sequential and parallel MPI programs Loading jobs and workflows into the Portal directly from the P-GRADE application development environment. Set up the list of Grids and define a set of default resources (portal administrator) and personalizing the setup of resources for Grids (users). Managing available Grid information systems (currently MDS2) to get the list of resources available in the Grid and retrieve detailed information about these resources. Transferring input and output files among storage and computational resources automatically. Executing workflows on NGS resources allocating jobs to available resources, Visualizing graphically, and in real time, the runtime execution of jobs and workflows, Monitoring NGS resources, Legacy code repository through GEMLCA (Grid Execution Management for Legacy Code Architecture). Monitoring P-GRADE Portal for Resource Availability Monitoring tools are necessary to ensure resource availability of a P-GRADE Portal. These tools enable GEMLCA system administrators to gauge system performance and classify the failed workflow components. This helps the user to identify and access working resources and execution of their workflows without getting entangled into failed workflow components. GEMLCA Monitoring Toolkit (GMT) is widely used a monitoring tool for GEMLCA Grid systems. GMT toolkit is based Globus MDS4 (Monitoring and Discovery System). It is used to automatically monitor and test computational grid and GEMLCA legacy applications at a predefined interval of time, which ensures the availabilities of these legacy applications. GMT also works as a complement for existing production grid brokers such as LCG broker. The availability of a legacy application over a period of time can be predicted by analyzing the previous behavioral pattern of particular application or resource deployed in grid system. Grid client users can predict the nature of an application by analyzing the probability of an application to be functioning at a time and mapping the execution of workflow components. Following diagram shows the role of third party monitoring tool in a Grid system. GMT scans the grid system for active services and update MDS indexing service. It maintains a historical data base of Grid applications performance which helps in searching of services over the Grid. A classifier component runs query on historical database to find out the available services to execute workflow submitted by a Grid portal. Web Services Resource Framework (WSRF) provides the framework for Monitoring and Discovery Systems (MDS), which is a part of Globus Monitoring toolkit distribution. MDS is an easy to use extensive query tool which can be used to search Grid for running services and post new workflows. Its graphical interface can also be used to gather WSRF resource properties and run third party tools to gather information about available resources. MDS use same indexing service to integrate different monitoring solutions. Following SQL query can be used to authorize a user to access historical database in order to access grid resources. if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[Main]') and OBJECTPROPERTY(id, N'IsUserTable') = 1) drop table [dbo].[Main] GO CREATE TABLE [dbo].[Main] ( [Probability_ID] [int] NOT NULL , [Service_ID] [int] NOT NULL , [NetworkConn_ID] [int] NOT NULL , [Registration_ID] [int] NOT NULL , [User_ID] [int] NOT NULL , [File_Size] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL , [Site_Status] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL , [Alerts] [varchar] (40) COLLATE SQL_Latin1_General_CP1_CI_AS NULL , [Program] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL , [Ping] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL , [Start_Time] [datetime] NULL , [End_Time] [datetime] NULL , [Time_of_Start] [datetime] NULL ) ON [PRIMARY] GO Following SQL server code is used to find out running services in a network: if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[FK__Main__Service_ID__1EA559DF]') and OBJECTPROPERTY(id, N'IsForeignKey') = 1) ALTER TABLE [dbo].[Main] DROP CONSTRAINT FK__Main__Service_ID__1EA559DF GO if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[Services]') and OBJECTPROPERTY(id, N'IsUserTable') = 1) drop table [dbo].[Services] GO CREATE TABLE [dbo].[Services] ( [Service_ID] [int] NOT NULL , [Type] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL ) ON [PRIMARY] GO Following SQL query can be used to query compute server for probability of resource availability. if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[FK__Main__Probabilit__1DB135A6]') and OBJECTPROPERTY(id, N'IsForeignKey') = 1) ALTER TABLE [dbo].[Main] DROP CONSTRAINT FK__Main__Probabilit__1DB135A6 GO if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[Probability]') and OBJECTPROPERTY(id, N'IsUserTable') = 1) drop table [dbo].[Probability] GO CREATE TABLE [dbo].[Probability] ( [Probability_ID] [int] NOT NULL , [Type] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL ) ON [PRIMARY] GO The following SQL Script can be used to query user information over the Grid: if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[FK__Main__User_ID__2181C68A]') and OBJECTPROPERTY(id, N'IsForeignKey') = 1) ALTER TABLE [dbo].[Main] DROP CONSTRAINT FK__Main__User_ID__2181C68A GO if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[Users]') and OBJECTPROPERTY(id, N'IsUserTable') = 1) drop table [dbo].[Users] GO CREATE TABLE [dbo].[Users] ( [User_ID] [int] NOT NULL , [Name] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL , [Location] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL , [Address] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL ) ON [PRIMARY] GO Following SQL server code check for network connectivity: if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[FK__Main__NetworkCon__1F997E18]') and OBJECTPROPERTY(id, N'IsForeignKey') = 1) ALTER TABLE [dbo].[Main] DROP CONSTRAINT FK__Main__NetworkCon__1F997E18 GO if exists (select * from dbo.sysobjects where id = object_id(N'[dbo].[Network_Connectivity]') and OBJECTPROPERTY(id, N'IsUserTable') = 1) drop table [dbo].[Network_Connectivity] GO CREATE TABLE [dbo].[Network_Connectivity] ( [NetworkConn_ID] [int] NOT NULL , [Type] [varchar] (30) COLLATE SQL_Latin1_General_CP1_CI_AS NULL ) ON [PRIMARY] GO Conclusion: This paper discusses various aspects of Grid computing, GEMLCA and its advantages. Architecture of GEMLCA framework has been discussed in length. It analyses the probability of resource availability over a grid. GEMLCA Monitoring tool and their functioning in an grid environment is also highlighted. Reference: 1. Thierry Delaitre, Tamas Kiss, Ariel Goyeneche, Gabor Terstyanszky, Stephen Winter and Peter Kacsuk, GEMLCA: Running Legacy Code Applications as Grid Services Journal of Grid Computing (2005), June 2005 2. Luigi Bitonti, Tamas Kiss, Gabor Terstyanszky, Thierry Delaitre, Stephen Winter and Peter Kacsuk, Dynamic Testing of Legacy Code Resources on the Grid, Hungary, June 2005 3. Centre for Parallel Computing, University of Westminster website, http://www.cpc.wmin.ac.uk/gemlca/ 4. D. Kuebler and W. Eibach, BAdapting Legacy Applications as Web Services^, IBM Developer Works, http://www-106.ibm.com/developerworks/webservices/ 5. Y. Huang, I. Taylor and D. W. Walker, BWrapping Legacy Codes for Grid-Based Applications^, in Proceedings of the 17th International Parallel and Distributed Processing Symposium, workshop on Java for HPC, 22Y26 April 2003, Nice, France. ISBN 0-7695-1926-1. 6. T. Bodhuin and M. Tortorella, BUsing Grid Technologies for Web-enabling Legacy Systems^, in Proceedings of the Software Technology and Engineering Practice (STEP), Software Analysis and Maintenance: Practices, Tools, Interoperability workshop, Amsterdam, The Netherlands, September 19Y21, 2003. 7. B. Balis, M. Bubak and M. Wegiel, BA Solution for Adapting Legacy Code as Web Services^, in V. Getov and T. Kiellmann (eds.),Component Models and Systems for Grid Applications, Springer, pp. 57Y75, 2005. ISBN 0-387-23351-2. 8. D. Gannon, S. Krishnan, A. Slominski, G. Kandaswamy, L. Fang, BBuilding Applications from a Web Service based Component Architecture^, in V. Getov and T. Kiellmann (eds.), Component Models and Systems for Grid Applications, Springer, pp. 3Y17, 2005. ISBN 0-387-23351-2. 9. I. Foster, C. Keselman, J. Nick, S. Tuecke, BThe Physiology of the Grid: An Open Grid Services Architecture for Distributed Systems integration^, 2002, http://www.globus.org/research/papers/anatomy.pdf 10. The Globus Alliance Website, http://www.globus.org/. 11. The g-Lite Website, http://glite.web.cern.ch/glite/. 12. T. Delaitre, A. Goyeneche, T. Kiss and S.C. Winter, BPublishing and Executing Parallel Legacy Code using an OGSI Grid Service^, Conf. Proc. of the 2004 International Conference on Computational Science and its Applications, Technical Session on Grid Computing, ISBN 3-540-22056-9, pp 30Y36, May 2004, Assisi, Italy. 13. Cs. Nemeth, G. Dozsa, R. Lovas, P. Kacsuk, BThe P-GRADE Grid portal^, in Computational Science and Its Applications Y ICCSA 2004: International Conference, Assisi, Italy, 2004, pp. 10Y19, LNCS 3044. 14. J. Novotny, M. Russell, O. Wehrens: GridSphere, Ban Advanced Portal Framework^, Conf. Proc. of the 30th EUROMICRO Conference, Rennes, France, August 31stY September 3rd 2004. 15. Condor Team, University of WisconsinYMadison: Condor Version 6.4.7 Manual, http://www.cs.wisc.edu/condor/ manual/v6.4/. 16. P. Kacsuk, A. Goyeneche, T. Delaitre, T. Kiss, Z. Farkas, T. Boczko, BHigh-level Grid Application Environment to Use Legacy Codes as OGSA Grid Services^, Conf. Proc. Of the 5th IEEE/ACM International Workshop on Grid Computing, Pittsburgh, USA, pp. 428Y435, November 8, 2004. 17. GAMESS-UK, CCLRC, http://www.cse.clrc.ac.uk/qcg/gamess-uk/. 18. MultiBayes, University of Reading, http://www.ams.rdg.ac.uk/. 19. A. Gourgoulis, G. Terstyansky, P. Kacsuk, S.C. Winter, BCreating Scalable Traffic Simulation on Clusters^, PDP2004. Conference Proceedings of the 12th Euromicro Conference on Parallel, Distributed and Network based Processing, La Coruna, Spain, 11Y13th February, 2004. 20. P. Kacsuk, G. Dozsa, J. Kovacs, R. Lovas, N. Podhorszky, Z. Balaton and G. Gombas, BP-GRADE: A Grid Programing Environment^, Journal of Grid Computing, Vol. 1. No. 2, pp. 171Y197, 2004. 21. V. Welch, F. Siebenlist, I. Foster, J. Bresnahan, K. Czajkowski, J. Gawor, C. Kesselman, S. Meder, L. Pearlman, S. Tuecke, BSecurity for Grid Services^, in Twelfth International Symposium on High Performance Distributed Computing (HPDC-12), IEEE Press, June 2003. http://www.globus.org/Security/GSI3/GT3-Security-HPDC.pdf 22. J. Gawor, S. Meder, F. Siebenlist and Von Welch, BGT3 Grid Security Infrastructure Overview^, Draft March, 17, 2005, Draft 17th, March, 2005. 23. A. Ting, W. Caixia and X. Yong, BDynamic Grid Service Deployment^, 2004, http://www.comp.nus.edu.sg/wangxb/SMA5505-2004/xieyong-report1.pdf. 24. J. B. Weissman, S. Kim and D. England, BSupporting the Dynamic Grid Service Lifecycle^, 2004, https://wwws.cs. umn.edu/tech_reports_upload/tr2004/04-041.pdf. 25. K. Czajkowski, S. Fitzgerald, I. Foster and C. Kesselman, BGrid Information Services for Distributed Resource Sharing^, in 10th IEEE International Symposium on High-Performance Distributed Computing (HPDC-10), IEEE Press, 2001. 26. J. Yu, S. Venugopal and R. Buyya, BGrid Market Directory: A Web Services-based Grid Service Publication Directory^, Technical Report, Grid Computing and Distributed Systems Lab, University of Melbourne, Jan 2003. 27. R. Al-Ali, A. Hafid, O.F. Rama and D.W. Walker, BQoS Adaptation in Service-Oriented Grids^, in Proceedings of the 1st International Workshop on Middleware for Grid Computing (MGC2003) at ACM/IFIP/USENIX Middleware 2003. Rio de Janeiro, Brazil, June 2003. 28. C. Dumitrescu and I. Foster: BGRUBER: A Grid Resource SLA Broker^, GriPhyN/iVDGL Technical Report, 2005. 29. S. Venugopal, R. Buyya and L. Winton, BA Grid Service Broker for Scheduling Distributed Data-Oriented Applications on Global Grids^, 2nd International Workshop on Middleware in Grid Computing, Toronto, Ontario, Canada, October 18, 2004. 30. R. Al-Ali, K. Amin and G. von Laszewski: BAn OGSA-based Quality of Service Framework^, 2nd International Workshop on Grid and Cooperative Computing (GCC2003), Shanghai, China, December 2003. 31. Simplified Wrapper and Interface Generator Y SWIG, Website http://www.swig.org. 32. The Triana Project Website Y http://www.trianacode.org/. 33. Open Middleware Infrastructure Institute Y OMII, Website http://www.omii.ac.uk/. 34. D.W Chadwick and O. Otenko, BThe PERMIS X509 role based privilege management infrastructure^, Future Generation Computing System, Vol. 19, No. 2, pp. 277Y289, Elsevier, 2003. 35. The UK e-Science OGSA Testbed Website Y http://dsg.port.ac.uk/projects/ogsa-testbed/. 36. Peter Kacsuk and Gergely Sipos, Multi-Grid, Multi-User Workflows in the P-GRADE Grid Portal, Journal of Grid Computing (2006), June 2006 37. LHC Computing Grid: http://lcg.web.cern.ch/LCG/. 38. M. Ellert et al., BThe NorduGrid project: Using Globus toolkit for building Grid infrastructure,^ Nuclear Instruments and Methods (Section A) 502, pp. 407Y410, 2003. 39. Grid2003 Team, BThe Grid2003 production Grid: Principles and practice^, in Proc. of the 13th IEEE Intl. Symposium on High Performance Distributed Computing (HPDC-13), 2004. 40. V. Sunderam and J. Dongarra, BPVM: A framework for parallel distributed computing^, Concurrency: Practice and Experience, Vol. 2, No. 4, pp. 315Y339, 1990. 41. B. Coghlan, J. Walsh and D. O'Callaghan, BThe Grid-Ireland deployment architecture^, in Proc. Of European Grid Conference, Amsterdam, The Netherlands, pp. 354Y363, 2005. 42. R. Allan et al., BVirtual research environments: Sakai demonstrator^, in Proc. Of UK e-Science All Hands Meeting, 2005. 43. L. Childers, T. Disz, R. Olson, M.E. Papka, R. Stevens and T. Udeshi, BAccess Grid: Immersive group-to-group collaborative visualization,^ in Proc. of the 4th International Immersive Projection Technology Workshop, 2000. 44. W. Gantzsch, BSun Grid engine, towards creating a compute power Grid^, in Proc. Of the 1st International Symposium on Cluster Computing and the Grid (CCGrid '01), IEEE Computer Society, pp. 35Y6, 2001. 45. S. Zhou, BLSF: Load sharing in large-scale Heterogeneous distributed systems^, in Proc. Of Workshop on Cluster Computing, 1992. 46. E. Seidel, G. Allen, A. Merzky and J. Nabrzyski, BGridLab Y A Grid application toolkit and testbed^, Future Generation Computer Systems, Vol. 18, No. 8, pp. 1143Y1153, 2002. 47. R. Alfieri, R. Cecchini, V. Ciaschini, L. dellAgnello, A. Frohner, A. Gianoli, K. Lorentey, , and F. Spata. Voms, an authorization system for virtual organizations. http://infnforge.cnaf.infn.it/voms/VOMS-Santiago.pdf. 48. S. Andreozzi, S. Fantinel, D. Rebatto, L. Vaccarossa, and G. Tortone. A monitoring tool for a grid operation center. In CHEP 2003, La Jolla, California, March 2003. 49. Mario Antonioletti, Malcolm Atkinson, Rob Baxter, Andrew Borley, Neil P Chue, Hong, Brian Collins, Neil Hardman, Ally Hume, Alan Knox, Mike Jackson, Amy Krause, Simon Laws, James Magowan, Norman W Paton, Dave Pearson, Tom Sugden, Paul Watson, and Martin Westhead. The design and implementation of grid database services in ogsa-dai. Concurrency and Computation: Practice and Experience, 17:357-376, 2005. 50. David Baker, Mark Baker, Hong Ong, and Helen Xiang. Integration and operational monitoring tools for the emerging uk e-science grid infrastructure. In Proceedings of the UK e-Science All Hands Meeting (AHM 2004), East Midlands Conference Centre, Nottingham, 2004. 51. B. Balis, M. Bubak, and M. Wegiel. A solution for adapting legacy code as web services. In V. Getov and T. Kiellmann, editors, Component Models and Systems for Grid Applications, pages 57-75. Springer, 2005. ISBN 0-387-23351-2. 52. C. Baru, Moore R, A. Rajasekar, and M. Wan. The sdsc storage resource broker. In Proc.CASCON'98 Conference, November 1998. Read More