In order to make an object based system fault tolerant, objects are replicated. We discuss a novel object based locking (OBL) protocol to lock replicated objects by extending the traditional quorum based protocols to abstract methods. If a pair of methods op/sub 1/ and op/sub 2/ are compatible, the quorums for op/sub 1/ and op/sub 2/ do not intersect in the OBL protocol, even if op/sub 1/ or op/sub 2/ updates the object. We propose a version vector to identify which methods are performed on a replica. We discuss the OBL protocol where the replicas exchange the compatible methods by using the version vector.
{"title":"Quorum-based locking protocol for replicated objects","authors":"Katsuya Tanaka, H. Higaki, M. Takizawa","doi":"10.1109/DOA.1999.794008","DOIUrl":"https://doi.org/10.1109/DOA.1999.794008","url":null,"abstract":"In order to make an object based system fault tolerant, objects are replicated. We discuss a novel object based locking (OBL) protocol to lock replicated objects by extending the traditional quorum based protocols to abstract methods. If a pair of methods op/sub 1/ and op/sub 2/ are compatible, the quorums for op/sub 1/ and op/sub 2/ do not intersect in the OBL protocol, even if op/sub 1/ or op/sub 2/ updates the object. We propose a version vector to identify which methods are performed on a replica. We discuss the OBL protocol where the replicas exchange the compatible methods by using the version vector.","PeriodicalId":360176,"journal":{"name":"Proceedings of the International Symposium on Distributed Objects and Applications","volume":"38 6","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133041506","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The paper describes work in progress and preliminary results from the ESPRIT project COMPASS (ESPRIT 25717) whose primary purpose is "showing the way" to the development of interoperable accounting software. An important part of the project work is to contribute to the standardisation of the General Ledger Facility within the Object Management Group (OMG). In this effort the COMPASS project and associated partners have shown the benefits of adopting RM-ODP principles when defining domain standards. The paper also describes the ongoing work in the project to validate this approach by producing pilot implementations of the new standard using both new and existing accounting software systems.
{"title":"Developing a global standard for interoperable accounting systems","authors":"R. Barnett, Tor Neple, J. Hassall","doi":"10.1109/DOA.1999.794019","DOIUrl":"https://doi.org/10.1109/DOA.1999.794019","url":null,"abstract":"The paper describes work in progress and preliminary results from the ESPRIT project COMPASS (ESPRIT 25717) whose primary purpose is \"showing the way\" to the development of interoperable accounting software. An important part of the project work is to contribute to the standardisation of the General Ledger Facility within the Object Management Group (OMG). In this effort the COMPASS project and associated partners have shown the benefits of adopting RM-ODP principles when defining domain standards. The paper also describes the ongoing work in the project to validate this approach by producing pilot implementations of the new standard using both new and existing accounting software systems.","PeriodicalId":360176,"journal":{"name":"Proceedings of the International Symposium on Distributed Objects and Applications","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134022480","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The Java/CORBA paradigm extends Java distributively and allows one to specify and implement multiplatform distributed object architectures. Such multi-client/server architectures, extended with object database management systems (OODBMS) as the principal persistent service, will become object based architectures (or OAs). The analysis and the use of dedicated techniques is required to specify and implement OAs, forcing us to challenge the current design and development approaches. We introduce a model for such an architecture that is based on a semantic distinction between "functional" objects and "data" objects. Functional objects (FOs) are Java/CORBA (client) server distributed objects with well defined (standard) IDL interfaces, data objects (DOs) are FOs "arguments/results" objects persistently stored in the OODBMS. The model relies on constructing a DO-graph, that is used for simple "dataflow" centralized calling mechanism for FOs, and that is induced by correct typed composition of FOs. An example of concrete realization of such a model is the implementation of the Text Retrieval and Enrichment for Vital Information (TREVI) distributed system for the analysis and categorization of natural language English texts (ESPRIT Project EP23311).
{"title":"A model for Java/CORBA and OODBMS distributed architectures","authors":"L. Mazzuccelli","doi":"10.1109/DOA.1999.793988","DOIUrl":"https://doi.org/10.1109/DOA.1999.793988","url":null,"abstract":"The Java/CORBA paradigm extends Java distributively and allows one to specify and implement multiplatform distributed object architectures. Such multi-client/server architectures, extended with object database management systems (OODBMS) as the principal persistent service, will become object based architectures (or OAs). The analysis and the use of dedicated techniques is required to specify and implement OAs, forcing us to challenge the current design and development approaches. We introduce a model for such an architecture that is based on a semantic distinction between \"functional\" objects and \"data\" objects. Functional objects (FOs) are Java/CORBA (client) server distributed objects with well defined (standard) IDL interfaces, data objects (DOs) are FOs \"arguments/results\" objects persistently stored in the OODBMS. The model relies on constructing a DO-graph, that is used for simple \"dataflow\" centralized calling mechanism for FOs, and that is induced by correct typed composition of FOs. An example of concrete realization of such a model is the implementation of the Text Retrieval and Enrichment for Vital Information (TREVI) distributed system for the analysis and categorization of natural language English texts (ESPRIT Project EP23311).","PeriodicalId":360176,"journal":{"name":"Proceedings of the International Symposium on Distributed Objects and Applications","volume":"364 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115978266","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Software evolution is one of the problematic areas in software management. In a distributed environment it is harder to tackle this problem because the dispersal of software makes it difficult to control the change, as well as the propagation of the change to whoever that is using the evolving service. The paper presents a model to alleviate this problem by making different distributed service versions substitutable. The mechanism comprises a mediator that enables clients of an old-version service to successfully request onto an instance of a new-version service. The mediator considers functionality compatibility, rather than operation signature compatibility, when mediating the request. Thus instead of forcing change on the client side, this model allows flexible interoperability between different versions of client and server software. To support the model, existing distributed object architectures may require some extension to their type repositories to maintain mapping information necessary for the work of the mediator.
{"title":"Enabling flexible cross-version interoperability for distributed services","authors":"T. Senivongse","doi":"10.1109/DOA.1999.794027","DOIUrl":"https://doi.org/10.1109/DOA.1999.794027","url":null,"abstract":"Software evolution is one of the problematic areas in software management. In a distributed environment it is harder to tackle this problem because the dispersal of software makes it difficult to control the change, as well as the propagation of the change to whoever that is using the evolving service. The paper presents a model to alleviate this problem by making different distributed service versions substitutable. The mechanism comprises a mediator that enables clients of an old-version service to successfully request onto an instance of a new-version service. The mediator considers functionality compatibility, rather than operation signature compatibility, when mediating the request. Thus instead of forcing change on the client side, this model allows flexible interoperability between different versions of client and server software. To support the model, existing distributed object architectures may require some extension to their type repositories to maintain mapping information necessary for the work of the mediator.","PeriodicalId":360176,"journal":{"name":"Proceedings of the International Symposium on Distributed Objects and Applications","volume":"363 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122831797","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The StratOSphere system provides a framework for distributed objects written in Java, unifying mobile code and distributed programming systems by providing the basic entities: relocatable instances and methods, persistent repositories, and mobile execution state. Each StratOSphere host provides a repository to store object instances and methods in a persistent manner The repository is partitioned among different hosts, to distribute the storage of objects, and to provide different implementations of an object specification among particular hosts. Client applications visit relevant repositories to acquire specialized behavior from methods stored within the repository. At run-time these entities can be accessed and dispatched, providing a means of invoking an operation in a dynamic fashion, while still ensuring type safety and correctness. In addition to unifying externally-defined methods with compiled objects, and providing location transparency through mobility, StratOSphere further unifies different versions of an object, permitting successive implementations to coexist within the system. We discuss the architecture and implementation of StratOSphere, describing how the goals of unifying different aspects of distributed computing have influenced its design.
{"title":"StratOSphere: unification of code, data, location, scope, and mobility","authors":"Daniel Wu, D. Agrawal, A. E. Abbadi","doi":"10.1109/DOA.1999.793973","DOIUrl":"https://doi.org/10.1109/DOA.1999.793973","url":null,"abstract":"The StratOSphere system provides a framework for distributed objects written in Java, unifying mobile code and distributed programming systems by providing the basic entities: relocatable instances and methods, persistent repositories, and mobile execution state. Each StratOSphere host provides a repository to store object instances and methods in a persistent manner The repository is partitioned among different hosts, to distribute the storage of objects, and to provide different implementations of an object specification among particular hosts. Client applications visit relevant repositories to acquire specialized behavior from methods stored within the repository. At run-time these entities can be accessed and dispatched, providing a means of invoking an operation in a dynamic fashion, while still ensuring type safety and correctness. In addition to unifying externally-defined methods with compiled objects, and providing location transparency through mobility, StratOSphere further unifies different versions of an object, permitting successive implementations to coexist within the system. We discuss the architecture and implementation of StratOSphere, describing how the goals of unifying different aspects of distributed computing have influenced its design.","PeriodicalId":360176,"journal":{"name":"Proceedings of the International Symposium on Distributed Objects and Applications","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121973238","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Electronic commerce on the Internet is evolving from simple customer-to-business interactions, like online shopping, to complex business-to-business extranet applications. These applications typically require back-office processing in two or more organisations. CORBA provides abstractions that make it a good technology for building such applications. Transactions are a well known technique for ensuring the overall consistency of system state in the presence of the concurrent access and occasional failure that are likely in extranet applications. The use of CORBA transactions for supporting extranet applications is complicated by the use of firewalls. Conventional firewall technology operates by restricting access based on host address and port number, and does not suit CORBA, which abstracts away from these concepts. The paper describes the issues involved and shows how they can be addressed using an advanced CORBA object gateway.
{"title":"CORBA transactions through firewalls","authors":"D. Ingham, Owen Rees, A. Norman","doi":"10.1109/DOA.1999.794046","DOIUrl":"https://doi.org/10.1109/DOA.1999.794046","url":null,"abstract":"Electronic commerce on the Internet is evolving from simple customer-to-business interactions, like online shopping, to complex business-to-business extranet applications. These applications typically require back-office processing in two or more organisations. CORBA provides abstractions that make it a good technology for building such applications. Transactions are a well known technique for ensuring the overall consistency of system state in the presence of the concurrent access and occasional failure that are likely in extranet applications. The use of CORBA transactions for supporting extranet applications is complicated by the use of firewalls. Conventional firewall technology operates by restricting access based on host address and port number, and does not suit CORBA, which abstracts away from these concepts. The paper describes the issues involved and shows how they can be addressed using an advanced CORBA object gateway.","PeriodicalId":360176,"journal":{"name":"Proceedings of the International Symposium on Distributed Objects and Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125723922","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
P. Chung, Adam Woei-Jyh Lee, Joanne Shih, S. Yajnik, Yennun Huang
Discusses experiments to study the behavior of distributed objects in the presence of failures. The work is motivated by a practical need in designing object-based distributed systems. System developers need to understand how objects fail and how to handle these failures in their design. We consider two distributed object platforms-DCOM and IONA's Orbix, an implementation of CORBA. We investigate nine potential failure scenarios. These correspond to three different failure types (hanging, abnormal termination and crashes) of three system components (threads, processes and machines). We design experiments to inject failures into server object executions. The results are presented as perceived by clients when these failures occur in the server objects. We apply the results of these experiments to evaluate the effectiveness of a set of simple monitoring and recovery mechanisms and also to suggest improvements in the current DCOM and Orbix implementations.
{"title":"Fault-injection experiments for distributed objects","authors":"P. Chung, Adam Woei-Jyh Lee, Joanne Shih, S. Yajnik, Yennun Huang","doi":"10.1109/DOA.1999.793991","DOIUrl":"https://doi.org/10.1109/DOA.1999.793991","url":null,"abstract":"Discusses experiments to study the behavior of distributed objects in the presence of failures. The work is motivated by a practical need in designing object-based distributed systems. System developers need to understand how objects fail and how to handle these failures in their design. We consider two distributed object platforms-DCOM and IONA's Orbix, an implementation of CORBA. We investigate nine potential failure scenarios. These correspond to three different failure types (hanging, abnormal termination and crashes) of three system components (threads, processes and machines). We design experiments to inject failures into server object executions. The results are presented as perceived by clients when these failures occur in the server objects. We apply the results of these experiments to evaluate the effectiveness of a set of simple monitoring and recovery mechanisms and also to suggest improvements in the current DCOM and Orbix implementations.","PeriodicalId":360176,"journal":{"name":"Proceedings of the International Symposium on Distributed Objects and Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1999-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128824381","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：