Pub Date : 1996-03-01DOI: 10.1088/0967-1846/3/1/008
E. Born
From a project on upgrading an existing commercially used operating system for support of a distributed system architecture, we present results and experience from a comprehensive analytical performance evaluation of a large number of different implementation strategies for distributed lock management. The various lock management policies we consider differ in the management of the lock database, in optimization concepts applied, and in the corresponding communication protocol. The system architectures discussed include a centralized lock management, a lock management with a partitioned lock database both with immediate and direct communication and with a token ring based communication, and a lock management with a replicated lock database and token ring communication. We derive analytical formulae for the dependence of expected response time of lock requests and their throughput, and provide a practical example.
{"title":"Analytical performance modelling of lock management in distributed systems","authors":"E. Born","doi":"10.1088/0967-1846/3/1/008","DOIUrl":"https://doi.org/10.1088/0967-1846/3/1/008","url":null,"abstract":"From a project on upgrading an existing commercially used operating system for support of a distributed system architecture, we present results and experience from a comprehensive analytical performance evaluation of a large number of different implementation strategies for distributed lock management. The various lock management policies we consider differ in the management of the lock database, in optimization concepts applied, and in the corresponding communication protocol. The system architectures discussed include a centralized lock management, a lock management with a partitioned lock database both with immediate and direct communication and with a token ring based communication, and a lock management with a replicated lock database and token ring communication. We derive analytical formulae for the dependence of expected response time of lock requests and their throughput, and provide a practical example.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130430967","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 : 1996-03-01DOI: 10.1088/0967-1846/3/1/007
G. Heijenk, B. Haverkort
The Asynchronous Transfer Mode (ATM) has been developed as a connection-oriented technique for the transfer of fixed-size cells over high-speed networks. Many applications, however, require a connectionless network service. In order to provide such a technique, one can built a connectionless service on top of the connection-oriented service. In doing so, the issue of connection management comes into play. In this paper we propose a new connection management mechanism that provides for low bandwidth usage (as compared to a permanent connection) and low delays (as compared to a connection-per-packet approach). We model the new mechanism under two workload scenarios: an ordinary Poisson process and an interrupted Poisson process. We use Markovian techniques as well as matrix-geometric methods to evaluate the new connection management mechanism. From the evaluations it turns out that the proposed mechanism is superior to older approaches (which can be seen as limiting cases).
{"title":"Design and evaluation of a connection management mechanism for an ATM-based connectionless service","authors":"G. Heijenk, B. Haverkort","doi":"10.1088/0967-1846/3/1/007","DOIUrl":"https://doi.org/10.1088/0967-1846/3/1/007","url":null,"abstract":"The Asynchronous Transfer Mode (ATM) has been developed as a connection-oriented technique for the transfer of fixed-size cells over high-speed networks. Many applications, however, require a connectionless network service. In order to provide such a technique, one can built a connectionless service on top of the connection-oriented service. In doing so, the issue of connection management comes into play. In this paper we propose a new connection management mechanism that provides for low bandwidth usage (as compared to a permanent connection) and low delays (as compared to a connection-per-packet approach). We model the new mechanism under two workload scenarios: an ordinary Poisson process and an interrupted Poisson process. We use Markovian techniques as well as matrix-geometric methods to evaluate the new connection management mechanism. From the evaluations it turns out that the proposed mechanism is superior to older approaches (which can be seen as limiting cases).","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122555276","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 : 1996-03-01DOI: 10.1088/0967-1846/3/1/004
Y. M. Teo, S. Tay
This paper presents an analytical model to evaluate the performance of parallel simulation on distributed computing platforms. The proposed model is formalized by two important time components in parallel and distributed processing: computation time and communication time. A conservative parallel simulation of multistage interconnection networks is used as an example in our analytical model. Performance metrics such as elapsed time, speedup and simulation bandwidth associated with different schemes for partitioning/mapping parallel simulation onto distributed processors are evaluated. Our mathematical analysis identifies the major constituents of simulation overheads in these mapping strategies necessary for improving parallel simulation efficiency. We also show that a perfectly balanced workload distribution may not necessarily translate into better performance. On the contrary, we have shown that a balanced mapping of workload may increase communication overheads resulting in a longer simulation elapsed time. Our performance model has been validated against implementation results from a parallel simulation model. The analytical framework is also practical to evaluate the runtime efficiency of other simulation applications which are based on the conservative paradigm.
{"title":"Performance analysis of parallel simulation on distributed systems","authors":"Y. M. Teo, S. Tay","doi":"10.1088/0967-1846/3/1/004","DOIUrl":"https://doi.org/10.1088/0967-1846/3/1/004","url":null,"abstract":"This paper presents an analytical model to evaluate the performance of parallel simulation on distributed computing platforms. The proposed model is formalized by two important time components in parallel and distributed processing: computation time and communication time. A conservative parallel simulation of multistage interconnection networks is used as an example in our analytical model. Performance metrics such as elapsed time, speedup and simulation bandwidth associated with different schemes for partitioning/mapping parallel simulation onto distributed processors are evaluated. Our mathematical analysis identifies the major constituents of simulation overheads in these mapping strategies necessary for improving parallel simulation efficiency. We also show that a perfectly balanced workload distribution may not necessarily translate into better performance. On the contrary, we have shown that a balanced mapping of workload may increase communication overheads resulting in a longer simulation elapsed time. Our performance model has been validated against implementation results from a parallel simulation model. The analytical framework is also practical to evaluate the runtime efficiency of other simulation applications which are based on the conservative paradigm.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"71 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125614983","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 : 1996-03-01DOI: 10.1088/0967-1846/3/1/002
C. Woodside, C. Schramm
The Layered System Generator is used to create synthetic distributed systems of tasks with client - server style (RPC) interactions, representing a wide range of software architectures and workload patterns. A synthetic task system can be used to generate network and workstation traffic which represents the load from a planned software system, so one can observe its probable performance when run on the target network, or its probable impact on other existing applications. It can be used to evaluate the planned software design, or the target network's capability, or both combined. Using LSG, tests were made with systems of up to 39 tasks on a UNIX network, to investigate the performance changes that occur when a small task system is scaled up in size. The performance recorded across the range of experiments was also compared with predictions made by an analytic performance model. The errors were found to be small provided an allowance is made for workstation daemons and similar load components.
{"title":"Scalability and performance experiments using synthetic distributed server systems","authors":"C. Woodside, C. Schramm","doi":"10.1088/0967-1846/3/1/002","DOIUrl":"https://doi.org/10.1088/0967-1846/3/1/002","url":null,"abstract":"The Layered System Generator is used to create synthetic distributed systems of tasks with client - server style (RPC) interactions, representing a wide range of software architectures and workload patterns. A synthetic task system can be used to generate network and workstation traffic which represents the load from a planned software system, so one can observe its probable performance when run on the target network, or its probable impact on other existing applications. It can be used to evaluate the planned software design, or the target network's capability, or both combined. Using LSG, tests were made with systems of up to 39 tasks on a UNIX network, to investigate the performance changes that occur when a small task system is scaled up in size. The performance recorded across the range of experiments was also compared with predictions made by an analytic performance model. The errors were found to be small provided an allowance is made for workstation daemons and similar load components.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"28 7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115707495","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 : 1996-03-01DOI: 10.1088/0967-1846/3/4/006
L. Russell, O. Wolfson, Clement T. Yu
To meet the demands of commercial data traffic on the information highway, a new look at managing data is necessary. One projected activity, sharing of point-of-sale information, is being considered in the Demand Activated Manufacturing Project of the American Textile Partnership project. A scenario is examined in which 100,000 retail outlets communicate over a period of days. They provide the latest estimate of demand for sewn products across a chain of 26,000 suppliers through the use of bill-of-materials explosions at four levels of detail. A new paradign the information flow manager, is developed to handle this situation, including the case where members of the supply chain fail to communicate and go out of business. Techniques for approximation are introduced to keep estimates of demand as current as possible.
{"title":"Information flow in the DAMA project beyond database managers: information flow managers","authors":"L. Russell, O. Wolfson, Clement T. Yu","doi":"10.1088/0967-1846/3/4/006","DOIUrl":"https://doi.org/10.1088/0967-1846/3/4/006","url":null,"abstract":"To meet the demands of commercial data traffic on the information highway, a new look at managing data is necessary. One projected activity, sharing of point-of-sale information, is being considered in the Demand Activated Manufacturing Project of the American Textile Partnership project. A scenario is examined in which 100,000 retail outlets communicate over a period of days. They provide the latest estimate of demand for sewn products across a chain of 26,000 suppliers through the use of bill-of-materials explosions at four levels of detail. A new paradign the information flow manager, is developed to handle this situation, including the case where members of the supply chain fail to communicate and go out of business. Techniques for approximation are introduced to keep estimates of demand as current as possible.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121363135","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 : 1996-03-01DOI: 10.1088/0967-1846/3/1/003
Gregory D. Peterson, R. Chamberlain
The utilization of networked, shared, heterogeneous workstations as an inexpensive parallel computational platform is an appealing idea. However, most performance models for parallel computation are oriented towards the use of tightly-coupled, dedicated, homogeneous processors. We develop and validate an analytic performance modelling methodology for synchronous iterative algorithms executing on networked workstations. The model includes the effects of application load, background load, and processor heterogeneity. We use two applications, nonlinear optimization and discrete-event simulation, to validate the model. Various policies for the use of the workstations are considered and the optimal (or near-optimal) scheduling found. The performance modelling methodology provides significant help in addressing scheduling and similar issues in a shared resource environment.
{"title":"Parallel application performance in a shared resource environment","authors":"Gregory D. Peterson, R. Chamberlain","doi":"10.1088/0967-1846/3/1/003","DOIUrl":"https://doi.org/10.1088/0967-1846/3/1/003","url":null,"abstract":"The utilization of networked, shared, heterogeneous workstations as an inexpensive parallel computational platform is an appealing idea. However, most performance models for parallel computation are oriented towards the use of tightly-coupled, dedicated, homogeneous processors. We develop and validate an analytic performance modelling methodology for synchronous iterative algorithms executing on networked workstations. The model includes the effects of application load, background load, and processor heterogeneity. We use two applications, nonlinear optimization and discrete-event simulation, to validate the model. Various policies for the use of the workstations are considered and the optimal (or near-optimal) scheduling found. The performance modelling methodology provides significant help in addressing scheduling and similar issues in a shared resource environment.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129314747","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 : 1996-03-01DOI: 10.1088/0967-1846/3/1/006
Luai M. Malhis, W. Sanders, R. Schlichting
Multicast protocols that provide message ordering and delivery guarantees are becoming increasingly important in distributed system design. However, despite the large number of such protocols, little analytical work has been done concerning their performance, especially in the presence of message loss. This paper illustrates a method for determining the performability of group multicast protocols using stochastic activity networks, a stochastic extension to Petri nets, and reduced base model construction. In particular, we study the performability of one such protocol, called Psync, under a wide variety of workload and message loss probabilities. The specific focus is on measuring two quantities, the stabilization time - that is, the time required for messages to arrive at all hosts - and channel utilization. The analysis shows that Psync works well when message transmissions are frequent, but it exhibits extremely long message stabilization times when transmissions are infrequent and message losses occur. We use this information to suggest a modification to Psync that greatly reduces stabilization time in this situation. The results provide useful insights into the behaviour of Psync, as well as serving as a guide for evaluating the performability of other group multicast protocols.
{"title":"Numerical performability evaluation of a group multicast protocol","authors":"Luai M. Malhis, W. Sanders, R. Schlichting","doi":"10.1088/0967-1846/3/1/006","DOIUrl":"https://doi.org/10.1088/0967-1846/3/1/006","url":null,"abstract":"Multicast protocols that provide message ordering and delivery guarantees are becoming increasingly important in distributed system design. However, despite the large number of such protocols, little analytical work has been done concerning their performance, especially in the presence of message loss. This paper illustrates a method for determining the performability of group multicast protocols using stochastic activity networks, a stochastic extension to Petri nets, and reduced base model construction. In particular, we study the performability of one such protocol, called Psync, under a wide variety of workload and message loss probabilities. The specific focus is on measuring two quantities, the stabilization time - that is, the time required for messages to arrive at all hosts - and channel utilization. The analysis shows that Psync works well when message transmissions are frequent, but it exhibits extremely long message stabilization times when transmissions are infrequent and message losses occur. We use this information to suggest a modification to Psync that greatly reduces stabilization time in this situation. The results provide useful insights into the behaviour of Psync, as well as serving as a guide for evaluating the performability of other group multicast protocols.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1996-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121597604","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 : 1995-12-01DOI: 10.1088/0967-1846/2/4/004
M. Kara
Distributed-controlled dynamic load balancing algorithms are known to have several advantages over centralized algorithms such as scalability, and fault tolerance. Distributed implies that the control is decentralized and that a copy of the algorithm (called a scheduler) is replicated on each host of the network. However, distributed control also contributes to the lack of global goals and lack of coherence. This paper presents a new algorithm called DGP (decentralized global plans) that addresses the problem of coherence and co-ordination in distributed dynamic load balancing algorithms. The DGP algorithm is based on a strategy called global plans (GP), and aims at maintaining all computational loads of a distributed system within a band called delta . The rationale for the design of DGP is to allow each scheduler to consider the actions of its peer schedulers. With this level of co-ordination, the schedulers can act more as a coherent team. This new approach first explicitly specifies a global goal and then designs a strategy around this global goal such that each scheduler (i) takes into account local decisions made by other schedulers; (ii) takes into account the effect of its local decisions on the overall system and (iii) ensures load balancing. An experimental evaluation of DGP with two other well known dynamic load balancing algorithms published in the literature shows that DGP performs consistently better. More significantly, the results indicate that the global plan approach provides a better framework for the design of distributed dynamic load balancing algorithms.
{"title":"A global plan policy for coherent co-operation in distributed dynamic load balancing algorithms","authors":"M. Kara","doi":"10.1088/0967-1846/2/4/004","DOIUrl":"https://doi.org/10.1088/0967-1846/2/4/004","url":null,"abstract":"Distributed-controlled dynamic load balancing algorithms are known to have several advantages over centralized algorithms such as scalability, and fault tolerance. Distributed implies that the control is decentralized and that a copy of the algorithm (called a scheduler) is replicated on each host of the network. However, distributed control also contributes to the lack of global goals and lack of coherence. This paper presents a new algorithm called DGP (decentralized global plans) that addresses the problem of coherence and co-ordination in distributed dynamic load balancing algorithms. The DGP algorithm is based on a strategy called global plans (GP), and aims at maintaining all computational loads of a distributed system within a band called delta . The rationale for the design of DGP is to allow each scheduler to consider the actions of its peer schedulers. With this level of co-ordination, the schedulers can act more as a coherent team. This new approach first explicitly specifies a global goal and then designs a strategy around this global goal such that each scheduler (i) takes into account local decisions made by other schedulers; (ii) takes into account the effect of its local decisions on the overall system and (iii) ensures load balancing. An experimental evaluation of DGP with two other well known dynamic load balancing algorithms published in the literature shows that DGP performs consistently better. More significantly, the results indicate that the global plan approach provides a better framework for the design of distributed dynamic load balancing algorithms.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123690869","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 : 1995-12-01DOI: 10.1088/0967-1846/2/4/001
J. Edwards, I. Coutts
Manufacturing processes are required to adapt to change as businesses respond to global competition. The paper describes a framework for building distributed manufacturing processes based on an integrating infrastructure. Through a decomposition based on application function, application interoperation and application interaction, proposals are made for structuring the integration software required to flexibly implement distributed systems which can evolve to support required change. The paper describes the architecture used in a proof of concept implementation of a machine vision process made up of distributed application objects. Support for change is demonstrated through comparison with a conventional distributed system which does not make use of the services of an integrating infrastructure. In describing the system architecture, requirements for `soft` CIM building blocks which can be `plugged` into an integrating infrastructure are defined. This work together with a range of approaches to application interoperation proposed by researchers at the MSI Research Institute are being further investigated through an EPSRC funded project entitled `Manufacturing Software Interoperability: Steps towards Interoperating Distributed Objects`.
{"title":"The flexible integration of machine objects within distributed manufacturing systems","authors":"J. Edwards, I. Coutts","doi":"10.1088/0967-1846/2/4/001","DOIUrl":"https://doi.org/10.1088/0967-1846/2/4/001","url":null,"abstract":"Manufacturing processes are required to adapt to change as businesses respond to global competition. The paper describes a framework for building distributed manufacturing processes based on an integrating infrastructure. Through a decomposition based on application function, application interoperation and application interaction, proposals are made for structuring the integration software required to flexibly implement distributed systems which can evolve to support required change. The paper describes the architecture used in a proof of concept implementation of a machine vision process made up of distributed application objects. Support for change is demonstrated through comparison with a conventional distributed system which does not make use of the services of an integrating infrastructure. In describing the system architecture, requirements for `soft` CIM building blocks which can be `plugged` into an integrating infrastructure are defined. This work together with a range of approaches to application interoperation proposed by researchers at the MSI Research Institute are being further investigated through an EPSRC funded project entitled `Manufacturing Software Interoperability: Steps towards Interoperating Distributed Objects`.","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122687382","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 : 1995-12-01DOI: 10.1088/0967-1846/2/4/005
A. Kshemkalyani, M. Raynal, M. Singhal
Recording on-the-fly global states of distributed executions is an important paradigm when one is interested in analysing, testing, or verifying properties associated with these executions. Since Chandy and Lamport`s (1985) seminal paper on this topic, this problem is called the snapshot problem. Unfortunately, the lack of both a globally shared memory and a global clock in a distributed system, added to the fact that transfer delays in these systems are finite but unpredictable, makes this problem non-trivial. This paper first discusses issues which have to be addressed to compute distributed snapshots in a consistent way. Then several algorithms which determine on-the-fly such snapshots are presented for several types of networks (according to the properties of their communication channels, namely, FIFO, non-FIFO, and causal delivery).
{"title":"An introduction to snapshot algorithms in distributed computing","authors":"A. Kshemkalyani, M. Raynal, M. Singhal","doi":"10.1088/0967-1846/2/4/005","DOIUrl":"https://doi.org/10.1088/0967-1846/2/4/005","url":null,"abstract":"Recording on-the-fly global states of distributed executions is an important paradigm when one is interested in analysing, testing, or verifying properties associated with these executions. Since Chandy and Lamport`s (1985) seminal paper on this topic, this problem is called the snapshot problem. Unfortunately, the lack of both a globally shared memory and a global clock in a distributed system, added to the fact that transfer delays in these systems are finite but unpredictable, makes this problem non-trivial. This paper first discusses issues which have to be addressed to compute distributed snapshots in a consistent way. Then several algorithms which determine on-the-fly such snapshots are presented for several types of networks (according to the properties of their communication channels, namely, FIFO, non-FIFO, and causal delivery).","PeriodicalId":404872,"journal":{"name":"Distributed Syst. Eng.","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1995-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128858957","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：