Pub Date : 1989-06-05DOI: 10.1109/ICDCS.1989.37929
Jonathan M. Smith, Gerald Q. Maguire
The task of concurrently computing alternative solutions to a problem where only one of the solutions is needed is examined. In this case the rule for selecting between the solutions is faster first, where the first successful alternative is selected. For problems where the required execution time is unpredictable, this method shows substantial execution time performance increases over other methods. In order to test the utility of the design, it is used for two application areas: distributed execution of recovery blocks and OR-parallelism in Prolog. The authors present: (1) a model for selection of alternatives in a sequential setting: (2) a transformation that allows alternatives to execute concurrently; (3) a description of the semantics-preservation mechanism; and (4) parameterization of where the performance improvements can be expected. Additionally, examples of application areas for the method are given.<>
{"title":"Transparent concurrent execution of mutually exclusive alternatives","authors":"Jonathan M. Smith, Gerald Q. Maguire","doi":"10.1109/ICDCS.1989.37929","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37929","url":null,"abstract":"The task of concurrently computing alternative solutions to a problem where only one of the solutions is needed is examined. In this case the rule for selecting between the solutions is faster first, where the first successful alternative is selected. For problems where the required execution time is unpredictable, this method shows substantial execution time performance increases over other methods. In order to test the utility of the design, it is used for two application areas: distributed execution of recovery blocks and OR-parallelism in Prolog. The authors present: (1) a model for selection of alternatives in a sequential setting: (2) a transformation that allows alternatives to execute concurrently; (3) a description of the semantics-preservation mechanism; and (4) parameterization of where the performance improvements can be expected. Additionally, examples of application areas for the method are given.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1989-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132480093","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 : 1989-06-05DOI: 10.1109/ICDCS.1989.37977
S. Sarin, R. Floyd, Nilkanth Phadnis
A replicated directory management algorithm is presented that supports full location-transparency by allowing migration and reconfiguration at the granularity of individual names. Initially, all names in a directory (including names not yet in use) are controlled by the same set of sites, but control over any individual name (or range of names) can be migrated, e.g. to the set of sites that control the data object associated with the name. The method is based on a directory replication algorithm of J.J. Bloch et al. (1987), incorporating two extensions: (1) allowing control over individual names in a directory to migrate independently; and (2) allowing the definition of relationships among names, such that if control over a given name migrates, then those names that depend on it are automatically migrated as well. Finally, it is stated that the major cost of the algorithm, compared to simpler approaches that do not provide the same degree of location transparency, is the additional complexity and difficulty of obtaining a consistent view of the entire directory.<>
{"title":"A flexible algorithm for replicated directory management","authors":"S. Sarin, R. Floyd, Nilkanth Phadnis","doi":"10.1109/ICDCS.1989.37977","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37977","url":null,"abstract":"A replicated directory management algorithm is presented that supports full location-transparency by allowing migration and reconfiguration at the granularity of individual names. Initially, all names in a directory (including names not yet in use) are controlled by the same set of sites, but control over any individual name (or range of names) can be migrated, e.g. to the set of sites that control the data object associated with the name. The method is based on a directory replication algorithm of J.J. Bloch et al. (1987), incorporating two extensions: (1) allowing control over individual names in a directory to migrate independently; and (2) allowing the definition of relationships among names, such that if control over a given name migrates, then those names that depend on it are automatically migrated as well. Finally, it is stated that the major cost of the algorithm, compared to simpler approaches that do not provide the same degree of location transparency, is the additional complexity and difficulty of obtaining a consistent view of the entire directory.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1989-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130160959","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 : 1989-06-05DOI: 10.1109/ICDCS.1989.37982
R. Kessler, M. Livny
Results obtained in a study of algorithms to implement a distributed-shared memory in a distributed (loosely coupled) environment are described. Distributed-shared memory is the implementation of shared memory across multiple nodes in a distributed system. This is accomplished using only the private memories of the nodes by controlling access to the pages of the shared memory and transferring data to and from the private memories when necessary. Alternative algorithms are analyzed to implement distributed-shared memory. The algorithms are analyzed and compared over a wide range of conditions. Application characteristics are identified which can be exploited by the algorithms. The conditions under which the algorithms analyzed perform better or worse than the other alternatives are shown. Results are obtained via simulation using a synthetic reference generator.<>
{"title":"An analysis of distributed shared memory algorithms","authors":"R. Kessler, M. Livny","doi":"10.1109/ICDCS.1989.37982","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37982","url":null,"abstract":"Results obtained in a study of algorithms to implement a distributed-shared memory in a distributed (loosely coupled) environment are described. Distributed-shared memory is the implementation of shared memory across multiple nodes in a distributed system. This is accomplished using only the private memories of the nodes by controlling access to the pages of the shared memory and transferring data to and from the private memories when necessary. Alternative algorithms are analyzed to implement distributed-shared memory. The algorithms are analyzed and compared over a wide range of conditions. Application characteristics are identified which can be exploited by the algorithms. The conditions under which the algorithms analyzed perform better or worse than the other alternatives are shown. Results are obtained via simulation using a synthetic reference generator.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1989-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132055089","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 : 1989-06-05DOI: 10.1109/ICDCS.1989.37978
D. Long, J. Carroll, K. Stewart
Several strategies for replica maintenance are considered, and the benefits of each are analyzed. Formulas describing the reliability of the replicated data object are presented, and closed-form solutions are given for the tractable cases. Numerical solutions, validated by simulation results, are used to analyze the tradeoffs between reliability and storage cost. With estimates of the mean times to site failure and repair in a given system, the numerical techniques presented can be applied to predict the fewest number of replicas required to provide the desired level of reliability.<>
{"title":"The reliability of regeneration-based replica control protocols","authors":"D. Long, J. Carroll, K. Stewart","doi":"10.1109/ICDCS.1989.37978","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37978","url":null,"abstract":"Several strategies for replica maintenance are considered, and the benefits of each are analyzed. Formulas describing the reliability of the replicated data object are presented, and closed-form solutions are given for the tractable cases. Numerical solutions, validated by simulation results, are used to analyze the tradeoffs between reliability and storage cost. With estimates of the mean times to site failure and repair in a given system, the numerical techniques presented can be applied to predict the fewest number of replicas required to provide the desired level of reliability.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1989-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131965202","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 : 1989-06-05DOI: 10.1109/ICDCS.1989.37981
Chia-Jiu Wang, V. Nelson, C. Wu
A message-passing computer architecture called the modified mesh-connected parallel computer (MMCPC) is proposed and studied. The MMCPC is designed to be general-purpose parallel architecture suitable for wafer-scale integration. Generalized stochastic Petri nets (GSPNs) are used to model the behavior of the MMCPC. The GSPN performance modeling results show a need for a new processing element (PE). A PE architecture, able to handle data processing and message passing concurrently, is proposed, and the silicon overhead is estimated in comparison with transputerlike PEs. Based on the proposed PE, optimum sizes of the MMCPC for different program structures are derived. A two-dimensional fast Fourier transform problem is used as an example to demonstrate that the MMCPC is a cost-effective performance-enhancement architecture to a real problem.<>
{"title":"Performance modeling of the modified mesh-connected parallel computer","authors":"Chia-Jiu Wang, V. Nelson, C. Wu","doi":"10.1109/ICDCS.1989.37981","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37981","url":null,"abstract":"A message-passing computer architecture called the modified mesh-connected parallel computer (MMCPC) is proposed and studied. The MMCPC is designed to be general-purpose parallel architecture suitable for wafer-scale integration. Generalized stochastic Petri nets (GSPNs) are used to model the behavior of the MMCPC. The GSPN performance modeling results show a need for a new processing element (PE). A PE architecture, able to handle data processing and message passing concurrently, is proposed, and the silicon overhead is estimated in comparison with transputerlike PEs. Based on the proposed PE, optimum sizes of the MMCPC for different program structures are derived. A two-dimensional fast Fourier transform problem is used as an example to demonstrate that the MMCPC is a cost-effective performance-enhancement architecture to a real problem.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1989-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133458362","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 : 1989-06-05DOI: 10.1109/ICDCS.1989.37951
A. Sheth
A simple model is used to study the effect of fault-tolerance techniques and system design on system availability. A generic multiprocessor architecture is used that can be configured in different ways to study the effect of system architectures. Important parameters studied are different system architectures and hardware fault-tolerance techniques, mean time to failure of basic components, database size and distribution, interconnect capacity, etc. Quantitative analysis compares the relative effect of different parameter values. Results show that the effect of different parameter values on system availability can be very significant. System architecture, use of hardware fault tolerance (particularly mirroring), and data storage methods emerge as very important parameters under the control of a system designer.<>
{"title":"Fault tolerance in a very large database system: a strawman analysis","authors":"A. Sheth","doi":"10.1109/ICDCS.1989.37951","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37951","url":null,"abstract":"A simple model is used to study the effect of fault-tolerance techniques and system design on system availability. A generic multiprocessor architecture is used that can be configured in different ways to study the effect of system architectures. Important parameters studied are different system architectures and hardware fault-tolerance techniques, mean time to failure of basic components, database size and distribution, interconnect capacity, etc. Quantitative analysis compares the relative effect of different parameter values. Results show that the effect of different parameter values on system availability can be very significant. System architecture, use of hardware fault tolerance (particularly mirroring), and data storage methods emerge as very important parameters under the control of a system designer.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1989-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128409168","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 : 1989-06-05DOI: 10.1109/ICDCS.1989.37971
A. Leff, Philip S. Yu, Yann-Hang Lee
The issue of transaction routing in a heterogeneous database environment is examined where transaction characteristics like reference locality implies that certain processing systems can be identified as being, in general, more suitable than others for a given transaction class. Routing which ignores these distinctions in an attempt to balance system load can degrade system performance. An adaptive routing strategy is considered which: (1) estimates the response time based on a steady-state analysis; (2) monitors how well actual response time conforms to the estimate; and (3) adaptively adjusts future estimates through a feedback control based on (2). It is found that the adaptive strategy greatly enhances performance robustness as compared to the results of the strategy without feedback. The feedback process used alleviates the sensitivity to accurate estimations of workload and system parameters. Various simulation studies are used to illustrate the adaptive strategy's robustness.<>
{"title":"Adaptive transaction routing in a heterogeneous database environment","authors":"A. Leff, Philip S. Yu, Yann-Hang Lee","doi":"10.1109/ICDCS.1989.37971","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37971","url":null,"abstract":"The issue of transaction routing in a heterogeneous database environment is examined where transaction characteristics like reference locality implies that certain processing systems can be identified as being, in general, more suitable than others for a given transaction class. Routing which ignores these distinctions in an attempt to balance system load can degrade system performance. An adaptive routing strategy is considered which: (1) estimates the response time based on a steady-state analysis; (2) monitors how well actual response time conforms to the estimate; and (3) adaptively adjusts future estimates through a feedback control based on (2). It is found that the adaptive strategy greatly enhances performance robustness as compared to the results of the strategy without feedback. The feedback process used alleviates the sensitivity to accurate estimations of workload and system parameters. Various simulation studies are used to illustrate the adaptive strategy's robustness.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1989-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116828060","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 : 1989-02-28DOI: 10.1109/ICDCS.1989.37959
R. Mirchandaney, D. Towsley, J. Stankovic
The performance characteristics of simple load-sharing algorithms are studied for heterogeneous distributed systems. It is assumed that non-negligible delays are encountered in transforming jobs from one node to another and in gathering remote state information. The effects of these delays on the performance of two algorithms called Forward and Reverse are analyzed. Queuing theoretic models are formulated for each of the algorithms operating in heterogeneous systems under the assumption that the job arrival process at each node is Poisson and the service times and job transfer time are exponentially distributed. The models are solved using the matrix-geometric solution technique. The models are tested with regard to the effects of varying thresholds, the impact of changing the probe limit, and the determination of the optimal response times over a large range of loads and delays. Wherever relevant, the results of the models are compared with M/M/1, random assignment, and the M/M/K models.<>
{"title":"Adaptive load sharing in heterogeneous systems","authors":"R. Mirchandaney, D. Towsley, J. Stankovic","doi":"10.1109/ICDCS.1989.37959","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37959","url":null,"abstract":"The performance characteristics of simple load-sharing algorithms are studied for heterogeneous distributed systems. It is assumed that non-negligible delays are encountered in transforming jobs from one node to another and in gathering remote state information. The effects of these delays on the performance of two algorithms called Forward and Reverse are analyzed. Queuing theoretic models are formulated for each of the algorithms operating in heterogeneous systems under the assumption that the job arrival process at each node is Poisson and the service times and job transfer time are exponentially distributed. The models are solved using the matrix-geometric solution technique. The models are tested with regard to the effects of varying thresholds, the impact of changing the probe limit, and the determination of the optimal response times over a large range of loads and delays. Wherever relevant, the results of the models are compared with M/M/1, random assignment, and the M/M/K models.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1989-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124294656","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 : 1900-01-01DOI: 10.1109/ICDCS.1989.37979
T. Ng, Shepherd S. B. Shi
A scheme to replicate transactions is described. The scheme allows a k-replicated transaction to survive (k-1) failures. No coordination among the k replicas is needed until one of them reaches the end and proceeds to abort the others. Consequently, the scheme avoids the overhead and delay caused by failure detection reconfiguration, and synchronization during the k replicas' execution. A robust commit protocol to choose the transaction replica that should be committed and a procedure to choose the nodes on which a transaction replica is executed are described. The goal of the procedure is to maximize reliability.<>
{"title":"Replicated transactions","authors":"T. Ng, Shepherd S. B. Shi","doi":"10.1109/ICDCS.1989.37979","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37979","url":null,"abstract":"A scheme to replicate transactions is described. The scheme allows a k-replicated transaction to survive (k-1) failures. No coordination among the k replicas is needed until one of them reaches the end and proceeds to abort the others. Consequently, the scheme avoids the overhead and delay caused by failure detection reconfiguration, and synchronization during the k replicas' execution. A robust commit protocol to choose the transaction replica that should be committed and a procedure to choose the nodes on which a transaction replica is executed are described. The goal of the procedure is to maximize reliability.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129385177","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 : 1900-01-01DOI: 10.1109/ICDCS.1989.37967
C. D. Wolfson, E. Voorhees, Maura M. Flatley
The authors describe a mechanism known as intelligent routers that provides a means to implement distributed applications in a heterogeneous environment. A router is a process that may explicitly request to migrate to another machine so it can avail itself of that machine's capabilities. The router environment supports the migration despite differences in the computer hardware, operating systems, and communication networks in the target set of computers. Three main goals are described for intelligent router systems: that the routers be self-directing; that the systems accommodate heterogeneity; and that the systems be robust. A description of the architecture of an intelligent router system is presented. Also presented is an implementation of a particular intelligent router system. This prototype supports simultaneous execution of multiple routers on a variety of hardware and operating systems; however, it is not yet robust. Finally, worm programs, Knos, and routers are compared and their differences are explored.<>
{"title":"Intelligent routers","authors":"C. D. Wolfson, E. Voorhees, Maura M. Flatley","doi":"10.1109/ICDCS.1989.37967","DOIUrl":"https://doi.org/10.1109/ICDCS.1989.37967","url":null,"abstract":"The authors describe a mechanism known as intelligent routers that provides a means to implement distributed applications in a heterogeneous environment. A router is a process that may explicitly request to migrate to another machine so it can avail itself of that machine's capabilities. The router environment supports the migration despite differences in the computer hardware, operating systems, and communication networks in the target set of computers. Three main goals are described for intelligent router systems: that the routers be self-directing; that the systems accommodate heterogeneity; and that the systems be robust. A description of the architecture of an intelligent router system is presented. Also presented is an implementation of a particular intelligent router system. This prototype supports simultaneous execution of multiple routers on a variety of hardware and operating systems; however, it is not yet robust. Finally, worm programs, Knos, and routers are compared and their differences are explored.<<ETX>>","PeriodicalId":266544,"journal":{"name":"[1989] Proceedings. The 9th International Conference on Distributed Computing Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134098247","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}