Pub Date : 1998-04-27DOI: 10.1088/0967-1846/5/4/006
J. Szymaszek, Andrzej Uszok, K. Zielinski
This paper presents experience gathered through the implementation of a CORBA-based localization system for an office environment. The localization system simultaneously preserves the fine-grained object-oriented structure of the system and achieves efficient performance. The presented study is a practical lesson concerning the implementation of a scalable, information-dissemination system. The key idea is to represent a large observable collection of objects by a repository that provides access to them both as individual CORBA objects and as data records. The proper use of this duality may have a substantial impact on the overall system performance. The repository is equipped with a scalable notification mechanism built around the concepts of a notification dispatcher and a notification tree. Fundamental features of the proposed solution are illustrated by a performance study and a representative application.
{"title":"Building a scalable and efficient component-oriented system using CORBA - an Active Badge system case study","authors":"J. Szymaszek, Andrzej Uszok, K. Zielinski","doi":"10.1088/0967-1846/5/4/006","DOIUrl":"https://doi.org/10.1088/0967-1846/5/4/006","url":null,"abstract":"This paper presents experience gathered through the implementation of a CORBA-based localization system for an office environment. The localization system simultaneously preserves the fine-grained object-oriented structure of the system and achieves efficient performance. The presented study is a practical lesson concerning the implementation of a scalable, information-dissemination system. The key idea is to represent a large observable collection of objects by a repository that provides access to them both as individual CORBA objects and as data records. The proper use of this duality may have a substantial impact on the overall system performance. The repository is equipped with a scalable notification mechanism built around the concepts of a notification dispatcher and a notification tree. Fundamental features of the proposed solution are illustrated by a performance study and a representative application.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115344236","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}
Pub Date : 1998-03-01DOI: 10.1088/0967-1846/5/1/002
M. Mowbray, G. Karlsson, Torsten Köhler
We introduce a new scheduling algorithm for multimedia traffic using capacity reservation. We compare it with other algorithms in the literature. It has been implemented and its worst-case performance has been analysed. It appears to give a noticeably improved quality of service to delay-sensitive traffic.
{"title":"Capacity reservation for multimedia traffic","authors":"M. Mowbray, G. Karlsson, Torsten Köhler","doi":"10.1088/0967-1846/5/1/002","DOIUrl":"https://doi.org/10.1088/0967-1846/5/1/002","url":null,"abstract":"We introduce a new scheduling algorithm for multimedia traffic using capacity reservation. We compare it with other algorithms in the literature. It has been implemented and its worst-case performance has been analysed. It appears to give a noticeably improved quality of service to delay-sensitive traffic.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128588991","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}
Pub Date : 1998-03-01DOI: 10.1088/0967-1846/5/1/003
H. Oliver, S. Brandt, Andrew Thomas, N. Charton
We describe our experiences and performance figures from an open-switching approach to broadband (ATM) connection management. The control system described is the lowest layer of a general connection management architecture such as TINA or Xbind.
{"title":"Network control as a distributed object application","authors":"H. Oliver, S. Brandt, Andrew Thomas, N. Charton","doi":"10.1088/0967-1846/5/1/003","DOIUrl":"https://doi.org/10.1088/0967-1846/5/1/003","url":null,"abstract":"We describe our experiences and performance figures from an open-switching approach to broadband (ATM) connection management. The control system described is the lowest layer of a general connection management architecture such as TINA or Xbind.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"192 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116145239","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}
Pub Date : 1998-03-01DOI: 10.1088/0967-1846/5/1/004
C. Karamanolis, J. Magee
This paper addresses the problem of constructing and managing highly available services for large, open distributed systems. A novel replication protocol is presented that satisfies two fundamental requirements of this environment. First, it hides replication from the service clients and secondly, it facilitates the dynamic reconfiguration of the server group. The protocol has been implemented and tested in the Regis distributed environment. The experimental results indicate that the cost of replication transparency is acceptable.
{"title":"Construction and management of highly available services in open distributed systems","authors":"C. Karamanolis, J. Magee","doi":"10.1088/0967-1846/5/1/004","DOIUrl":"https://doi.org/10.1088/0967-1846/5/1/004","url":null,"abstract":"This paper addresses the problem of constructing and managing highly available services for large, open distributed systems. A novel replication protocol is presented that satisfies two fundamental requirements of this environment. First, it hides replication from the service clients and secondly, it facilitates the dynamic reconfiguration of the server group. The protocol has been implemented and tested in the Regis distributed environment. The experimental results indicate that the cost of replication transparency is acceptable.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1998-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133965155","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}
Pub Date : 1997-12-01DOI: 10.1088/0967-1846/4/4/003
Patrick Martin, W. Powley
Multidatabase systems (MDBSs) provide applications with integrated access to a collection of databases. The component database systems are typically heterogeneous, distributed, autonomous and pre-existing. MDBSs, like conventional database management systems (DBMSs), require catalogue information to provide their services. The catalogue must be stored in a separate system database, or repository, since component databases are assumed to be independent of the MDBS. We examine the approach taken to catalogue management in the CORDS MDBS which uses an X.500 directory service to store the MDBS catalogue and discuss the advantages and disadvantages of this approach. Storing the MDBS catalogue in a directory service is an appealing approach but is an application for which the directory service was never intended. There are, therefore, practical problems which must be addressed including the suitability of the information model and the performance of the system for typical repository requests. We discuss the design of the CORDS MDBS catalogue and present a set of experiments we conducted with it to examine the performance and scalability of the approach.
{"title":"Catalogue management for a multidatabase system using an X.500 directory system","authors":"Patrick Martin, W. Powley","doi":"10.1088/0967-1846/4/4/003","DOIUrl":"https://doi.org/10.1088/0967-1846/4/4/003","url":null,"abstract":"Multidatabase systems (MDBSs) provide applications with integrated access to a collection of databases. The component database systems are typically heterogeneous, distributed, autonomous and pre-existing. MDBSs, like conventional database management systems (DBMSs), require catalogue information to provide their services. The catalogue must be stored in a separate system database, or repository, since component databases are assumed to be independent of the MDBS. We examine the approach taken to catalogue management in the CORDS MDBS which uses an X.500 directory service to store the MDBS catalogue and discuss the advantages and disadvantages of this approach. Storing the MDBS catalogue in a directory service is an appealing approach but is an application for which the directory service was never intended. There are, therefore, practical problems which must be addressed including the suitability of the information model and the performance of the system for typical repository requests. We discuss the design of the CORDS MDBS catalogue and present a set of experiments we conducted with it to examine the performance and scalability of the approach.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132260720","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}
Pub Date : 1997-12-01DOI: 10.1088/0967-1846/4/4/005
Xiaomin Chen, L. Moser, P. Melliar-Smith
Totally ordered multicast protocols are important in distributed system design, because totally ordered multicasts provide a considerably simpler programming paradigm than causally or unordered multicasts. These protocols convert the underlying unreliable and unordered message delivery provided by the network into high-quality reliable totally ordered message delivery. The emerging gigabit networks provide both ultra-high data rates and extremely low error rates. These characteristics strongly challenge the basic assumptions made for existing totally ordered multicast protocols. In this paper, we present the QuickRing multicast protocol, which provides totally ordered multicasting for the ring-based gigabit network QuickRing. The protocol is based on a buffer reservation strategy and uses timestamps both for buffer reservations and for total ordering of messages. Totally ordered multicasting implemented in hardware achieves low overhead, low latency, high throughput and low cost.
{"title":"Totally ordered gigabit multicasting","authors":"Xiaomin Chen, L. Moser, P. Melliar-Smith","doi":"10.1088/0967-1846/4/4/005","DOIUrl":"https://doi.org/10.1088/0967-1846/4/4/005","url":null,"abstract":"Totally ordered multicast protocols are important in distributed system design, because totally ordered multicasts provide a considerably simpler programming paradigm than causally or unordered multicasts. These protocols convert the underlying unreliable and unordered message delivery provided by the network into high-quality reliable totally ordered message delivery. The emerging gigabit networks provide both ultra-high data rates and extremely low error rates. These characteristics strongly challenge the basic assumptions made for existing totally ordered multicast protocols. In this paper, we present the QuickRing multicast protocol, which provides totally ordered multicasting for the ring-based gigabit network QuickRing. The protocol is based on a buffer reservation strategy and uses timestamps both for buffer reservations and for total ordering of messages. Totally ordered multicasting implemented in hardware achieves low overhead, low latency, high throughput and low cost.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"156 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115587444","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}
Pub Date : 1997-12-01DOI: 10.1088/0967-1846/4/4/006
R. Baldoni, S. Salza
Deadlock detection is an interesting problem in multidatabase systems (MDBS), since the autonomy of the local systems precludes the visibility of the state of local transactions and the contention on items, and therefore the classical approaches proposed for homogeneous distributed database systems cannot be extended to the MDBS case. A few specific methods have been proposed in the literature to detect potential global deadlocks, that do not necessarily correspond to real ones. In this paper we present a comparative performance study of several global deadlock detection methods. The results of the analysis have suggested a new hybrid deadlock detection method, that we present in the paper and that is very well suited for a distributed implementation and has a performance that, according to our experiments, compares favourably with all the other methods in a variety of workload conditions.
{"title":"Deadlock detection in multidatabase systems: a performance analysis","authors":"R. Baldoni, S. Salza","doi":"10.1088/0967-1846/4/4/006","DOIUrl":"https://doi.org/10.1088/0967-1846/4/4/006","url":null,"abstract":"Deadlock detection is an interesting problem in multidatabase systems (MDBS), since the autonomy of the local systems precludes the visibility of the state of local transactions and the contention on items, and therefore the classical approaches proposed for homogeneous distributed database systems cannot be extended to the MDBS case. A few specific methods have been proposed in the literature to detect potential global deadlocks, that do not necessarily correspond to real ones. In this paper we present a comparative performance study of several global deadlock detection methods. The results of the analysis have suggested a new hybrid deadlock detection method, that we present in the paper and that is very well suited for a distributed implementation and has a performance that, according to our experiments, compares favourably with all the other methods in a variety of workload conditions.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"124 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128109416","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}
Pub Date : 1997-12-01DOI: 10.1088/0967-1846/4/4/004
C. Logé, E. Najm, Ken Chen
A next generation of systems that should appear will have to manage simultaneously several geographically distributed users. These systems belong to the class of computer-supported cooperative work systems (CSCW). The development of such complex systems requires rigorous development methods and flexible open architectures. Open distributed processing (ODP) is a standardization effort that aims at providing such architectures. ODP features appropriate abstraction levels and a clear articulation between requirements, programming and infrastructure support. ODP advocates the use of formal methods for the specification of systems and components. The computational model, an object-based model, one of the abstraction levels identified within ODP, plays a central role in the global architecture. In this model, basic objects can be composed with communication and distribution abstractions (called binding objects) to form a computational specification of distributed systems, or applications. Computational specifications can then be mapped (in a mechanism akin to compilation) onto an engineering solution. We use an ODP-inspired method to computationally specify a cooperative system. We start from a general purpose component that we progressively refine into a collection of basic and binding objects. We focus on two issues of a co-authoring application, namely, dynamic reconfiguration and multiview synchronization. We discuss solutions for these issues and formalize them using the MT-LOTOS specification language that is currently studied in the ISO standardization formal description techniques group.
{"title":"An ODP computational model of a cooperative binding object","authors":"C. Logé, E. Najm, Ken Chen","doi":"10.1088/0967-1846/4/4/004","DOIUrl":"https://doi.org/10.1088/0967-1846/4/4/004","url":null,"abstract":"A next generation of systems that should appear will have to manage simultaneously several geographically distributed users. These systems belong to the class of computer-supported cooperative work systems (CSCW). The development of such complex systems requires rigorous development methods and flexible open architectures. Open distributed processing (ODP) is a standardization effort that aims at providing such architectures. ODP features appropriate abstraction levels and a clear articulation between requirements, programming and infrastructure support. ODP advocates the use of formal methods for the specification of systems and components. The computational model, an object-based model, one of the abstraction levels identified within ODP, plays a central role in the global architecture. In this model, basic objects can be composed with communication and distribution abstractions (called binding objects) to form a computational specification of distributed systems, or applications. Computational specifications can then be mapped (in a mechanism akin to compilation) onto an engineering solution. We use an ODP-inspired method to computationally specify a cooperative system. We start from a general purpose component that we progressively refine into a collection of basic and binding objects. We focus on two issues of a co-authoring application, namely, dynamic reconfiguration and multiview synchronization. We discuss solutions for these issues and formalize them using the MT-LOTOS specification language that is currently studied in the ISO standardization formal description techniques group.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131140094","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}
Pub Date : 1997-12-01DOI: 10.1088/0967-1846/4/4/002
Yen-Min Huang, C. Ravishankar
Protocol heterogeneity is pervasive and is a major obstacle to effective integration of services in large systems. However, standardization is not a complete answer. Standardized protocols must be general to prevent a proliferation of standards, and can therefore become complex and inefficient. Specialized protocols can be simple and efficient, since they can ignore situations that are precluded by application characteristics. One solution is to maintain agents for translating between protocols. However, n protocol types would require agents, since an agent must exist for a source - destination pair. A better solution is to create agents as needed. This paper examines the issues in the creation and management of protocol translation agents. We focus on the design of Nestor, an environment for synthesizing and managing RPC protocol translation agents. We provide rationale for the translation mechanism and the synthesis environment, with specific emphasis on the security issues arising in Nestor. Nestor has been implemented and manages heterogeneous RPC agents generated using the Cicero protocol construction language and the URPC toolkit.
{"title":"Secure synthesis and activation of protocol translation agents","authors":"Yen-Min Huang, C. Ravishankar","doi":"10.1088/0967-1846/4/4/002","DOIUrl":"https://doi.org/10.1088/0967-1846/4/4/002","url":null,"abstract":"Protocol heterogeneity is pervasive and is a major obstacle to effective integration of services in large systems. However, standardization is not a complete answer. Standardized protocols must be general to prevent a proliferation of standards, and can therefore become complex and inefficient. Specialized protocols can be simple and efficient, since they can ignore situations that are precluded by application characteristics. One solution is to maintain agents for translating between protocols. However, n protocol types would require agents, since an agent must exist for a source - destination pair. A better solution is to create agents as needed. This paper examines the issues in the creation and management of protocol translation agents. We focus on the design of Nestor, an environment for synthesizing and managing RPC protocol translation agents. We provide rationale for the translation mechanism and the synthesis environment, with specific emphasis on the security issues arising in Nestor. Nestor has been implemented and manages heterogeneous RPC agents generated using the Cicero protocol construction language and the URPC toolkit.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"114 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133614657","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}
Pub Date : 1997-12-01DOI: 10.1088/0967-1846/4/4/001
Wayne Brookes, S. Crawley, J. Indulska, Douglas Kosovic, A. Vogel
Open distributed processing aims to support cooperation within and between large-scale heterogeneous and autonomous computing environments. An inherent issue in such environments is enabling the interoperation of objects whose interfaces have been defined in different type models. In this paper, we present a type management system which provides a means for representing, storing, retrieving and translating types, and for expressing and evaluating relationships between types in a heterogeneous distributed computing environment. This system allows multiple type languages and models, and can relate types expressed in different ones. The type management system is designed to support the instantiation and dynamic binding of objects, run-time type checking of object interactions, and the discovery of new resources (e.g. services) within the system. Current approaches to interface definition in distributed systems are mainly based on the use of a single interface definition language (IDL). While this provides a level of common agreement about the types of system interfaces, the type models of existing IDLs are not rich enough to model either the overall architecture of a system or the behaviour of objects. We illustrate this by briefly describing some aspects of an enhanced type model with the emphasis on the model's impact on the design of the type management system.
{"title":"Types and their management in open distributed systems","authors":"Wayne Brookes, S. Crawley, J. Indulska, Douglas Kosovic, A. Vogel","doi":"10.1088/0967-1846/4/4/001","DOIUrl":"https://doi.org/10.1088/0967-1846/4/4/001","url":null,"abstract":"Open distributed processing aims to support cooperation within and between large-scale heterogeneous and autonomous computing environments. An inherent issue in such environments is enabling the interoperation of objects whose interfaces have been defined in different type models. In this paper, we present a type management system which provides a means for representing, storing, retrieving and translating types, and for expressing and evaluating relationships between types in a heterogeneous distributed computing environment. This system allows multiple type languages and models, and can relate types expressed in different ones. The type management system is designed to support the instantiation and dynamic binding of objects, run-time type checking of object interactions, and the discovery of new resources (e.g. services) within the system. Current approaches to interface definition in distributed systems are mainly based on the use of a single interface definition language (IDL). While this provides a level of common agreement about the types of system interfaces, the type models of existing IDLs are not rich enough to model either the overall architecture of a system or the behaviour of objects. We illustrate this by briefly describing some aspects of an enhanced type model with the emphasis on the model's impact on the design of the type management system.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133665394","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：