Pub Date : 1991-12-01DOI: 10.1109/PDIS.1991.183066
S. Su, Yaw-Huei Chen, H. Lam
The authors have show previously (1989, 1991) that processing OODBs can be viewed as the manipulation of patterns of object associations. Parallel, multiple wavefront algorithms based on an identification approach for verifying association patterns have been introduced. The current paper presents an elimination approach for processing OODBs. The new approach allows more processors to operate concurrently on a query, thus allowing a higher degree of parallelism in query processing. A formal proof of the correctness of the new approach is given, and a parallel elimination algorithm for processing tree queries is presented. Some simulation results are also provided to compare the performance of the identification approach with the elimination approach.<>
{"title":"Multiple wavefront algorithms for pattern-based processing of object-oriented databases","authors":"S. Su, Yaw-Huei Chen, H. Lam","doi":"10.1109/PDIS.1991.183066","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183066","url":null,"abstract":"The authors have show previously (1989, 1991) that processing OODBs can be viewed as the manipulation of patterns of object associations. Parallel, multiple wavefront algorithms based on an identification approach for verifying association patterns have been introduced. The current paper presents an elimination approach for processing OODBs. The new approach allows more processors to operate concurrently on a query, thus allowing a higher degree of parallelism in query processing. A formal proof of the correctness of the new approach is given, and a parallel elimination algorithm for processing tree queries is presented. Some simulation results are also provided to compare the performance of the identification approach with the elimination approach.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114980458","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183114
B. Dageville, Pascale Borla-Salamet, C. Chachaty
Parallel execution of database queries in a shared nothing multiprocessor environment, needs control mechanisms which guarantee a correct global execution of each query. In the authors system, the control is integrated into the query code at compile time. According to the required schedule for the query, the compilation process combines a set of basic control operations into a control graph mapped on the query graph. This control which allows pipelining between data operations, achieves the dataflow control and enforces the schedule of program. Using fine grain control operations combined into control schemas makes the control flexible, as new strategies could be easily supported. In their prototype the parallel LERA language is used.<>
{"title":"Compiling control into database queries for parallel execution management","authors":"B. Dageville, Pascale Borla-Salamet, C. Chachaty","doi":"10.1109/PDIS.1991.183114","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183114","url":null,"abstract":"Parallel execution of database queries in a shared nothing multiprocessor environment, needs control mechanisms which guarantee a correct global execution of each query. In the authors system, the control is integrated into the query code at compile time. According to the required schedule for the query, the compilation process combines a set of basic control operations into a control graph mapped on the query graph. This control which allows pipelining between data operations, achieves the dataflow control and enforces the schedule of program. Using fine grain control operations combined into control schemas makes the control flexible, as new strategies could be easily supported. In their prototype the parallel LERA language is used.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123069869","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183098
R. Dasgupta, S. Bhattacharya
Summary form only given. The authors propose an efficient data organization for the storage and subsequent retrieval of large volume of cartographic data. The concept of address squares has been used extensively. The inherent parallelism in the data has been explored in order to improve computational efficiency. Several parallel algorithms related to the analysis of the cartographic data based on the proposed data organisation, have also been presented.<>
{"title":"An efficient data organization for parallel algorithmic design of cartographic database","authors":"R. Dasgupta, S. Bhattacharya","doi":"10.1109/PDIS.1991.183098","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183098","url":null,"abstract":"Summary form only given. The authors propose an efficient data organization for the storage and subsequent retrieval of large volume of cartographic data. The concept of address squares has been used extensively. The inherent parallelism in the data has been explored in order to improve computational efficiency. Several parallel algorithms related to the analysis of the cartographic data based on the proposed data organisation, have also been presented.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130982926","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183085
Z. Lin
An execution model based on concurrent frame signature file (CAT) is proposed for full text retrieval. The proposed method places emphasis on creating opportunities for concurrent access to the signature file, which resides on disks, without resorting to redundant storage. Thus, the advantage of low space overhead availed by signature file methods is retained, while the response time of retrieval is significantly improved when implemented on a multiprocessor with multiple disks. Performance of the proposed method is measured by analysis and experiment. Results from a simulation study on the BBN Butterfly TC2000 multiprocessor shows the speed of the proposed method is very promising in both relative and absolute terms.<>
{"title":"CAT: an execution model for concurrent full text search","authors":"Z. Lin","doi":"10.1109/PDIS.1991.183085","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183085","url":null,"abstract":"An execution model based on concurrent frame signature file (CAT) is proposed for full text retrieval. The proposed method places emphasis on creating opportunities for concurrent access to the signature file, which resides on disks, without resorting to redundant storage. Thus, the advantage of low space overhead availed by signature file methods is retained, while the response time of retrieval is significantly improved when implemented on a multiprocessor with multiple disks. Performance of the proposed method is measured by analysis and experiment. Results from a simulation study on the BBN Butterfly TC2000 multiprocessor shows the speed of the proposed method is very promising in both relative and absolute terms.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124492173","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183101
D. Kotz, C. Ellis
Improvements in the processing speed of multiprocessors are outpacing improvements in the speed of disk hardware. Parallel disk I/O subsystems have been proposed as one way to close the gap between processor and disk speeds. In a previous paper the authors showed that prefetching and caching have the potential to deliver the performance benefits of parallel file systems to parallel applications. They describe experiments with practical prefetching policies, and show that prefetching can be implemented efficiently even for the more complex parallel file access patterns. They also test the ability of these policies across a range of architectural parameters. (see IEEE Trans. on Parallel and Distributed Systems, vol.1, no.2, p.218-30, 1990).<>
{"title":"Practical prefetching techniques for parallel file systems","authors":"D. Kotz, C. Ellis","doi":"10.1109/PDIS.1991.183101","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183101","url":null,"abstract":"Improvements in the processing speed of multiprocessors are outpacing improvements in the speed of disk hardware. Parallel disk I/O subsystems have been proposed as one way to close the gap between processor and disk speeds. In a previous paper the authors showed that prefetching and caching have the potential to deliver the performance benefits of parallel file systems to parallel applications. They describe experiments with practical prefetching policies, and show that prefetching can be implemented efficiently even for the more complex parallel file access patterns. They also test the ability of these policies across a range of architectural parameters. (see IEEE Trans. on Parallel and Distributed Systems, vol.1, no.2, p.218-30, 1990).<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124612891","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183080
M. Houtsma, Filippo Cacace, S. Ceri
Presents a new approach to parallel computation of transitive closure queries using a semantic data fragmentation. Tuples of a large base relation denote edges in a graph, which models a transportation network. A fragmentation algorithm is proposed which produces a partitioning of the base relation into several fragments such that any fragment corresponds to a subgraph. One fragment, called high-speed fragment, collects all edges which guarantee maximum speed. Thus, the fragmentation algorithm induces a hierarchical relationship between the high-speed fragment and all other fragments. With this fragmentation, any query about paths connecting two nodes can be answered by using just the fragments in which nodes are located and the high-speed fragment. In general, if each fragment is managed by a distinguished processor, then the query can be answered by three processors working in parallel. This schema can be applied recursively to generate an arbitrary number of hierarchical levels.<>
{"title":"Parallel hierarchical evaluation of transitive closure queries","authors":"M. Houtsma, Filippo Cacace, S. Ceri","doi":"10.1109/PDIS.1991.183080","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183080","url":null,"abstract":"Presents a new approach to parallel computation of transitive closure queries using a semantic data fragmentation. Tuples of a large base relation denote edges in a graph, which models a transportation network. A fragmentation algorithm is proposed which produces a partitioning of the base relation into several fragments such that any fragment corresponds to a subgraph. One fragment, called high-speed fragment, collects all edges which guarantee maximum speed. Thus, the fragmentation algorithm induces a hierarchical relationship between the high-speed fragment and all other fragments. With this fragmentation, any query about paths connecting two nodes can be answered by using just the fragments in which nodes are located and the high-speed fragment. In general, if each fragment is managed by a distinguished processor, then the query can be answered by three processors working in parallel. This schema can be applied recursively to generate an arbitrary number of hierarchical levels.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127064286","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183072
Jürgen Seib
A Datalog program consists of function-free Horn clause rules. There are, however, some situations described more easily and more naturally by the use of general terms as arguments. A term is built up from variables, constants, and function symbols. A set of rules where the rules may contain function symbols is called a Datalog/sup Fun/ program. The paper discusses parallel processing of decomposable Datalog/sup Fun/ programs to overcome the performance problem. A decomposable program can be evaluated in parallel such that neither a communication nor a synchronization of the processors has to be established. The author and G. Lausen (1991) proposed the concept of generalized pivoting as a sufficient condition for decomposability of arbitrary but function-free Datalog programs. The current paper extends the concept of generalized pivoting to programs which may contain function symbols.<>
{"title":"Parallelizing datalog: function symbols make a difference","authors":"Jürgen Seib","doi":"10.1109/PDIS.1991.183072","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183072","url":null,"abstract":"A Datalog program consists of function-free Horn clause rules. There are, however, some situations described more easily and more naturally by the use of general terms as arguments. A term is built up from variables, constants, and function symbols. A set of rules where the rules may contain function symbols is called a Datalog/sup Fun/ program. The paper discusses parallel processing of decomposable Datalog/sup Fun/ programs to overcome the performance problem. A decomposable program can be evaluated in parallel such that neither a communication nor a synchronization of the processors has to be established. The author and G. Lausen (1991) proposed the concept of generalized pivoting as a sufficient condition for decomposability of arbitrary but function-free Datalog programs. The current paper extends the concept of generalized pivoting to programs which may contain function symbols.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115379582","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183115
D. DeWitt, J. Naughton, Donovan A. Schneider
The authors consider the problem of external sorting in a shared-nothing multiprocessor. A critical step in the algorithms the authors consider is to determine the range of sort keys to be handled by each processor. They consider two techniques for determining these ranges of sort keys: exact splitting, using a parallel version of the algorithm proposed by Iyer, Ricard, and Varman; and probabilistic splitting, which uses sampling to estimate quantiles. They present analytic results showing that probabilistic splitting performs better than exact splitting. Finally, the authors present experimental results from an implementation of sorting probabilistic splitting in the Gamma parallel database machine.<>
{"title":"Parallel sorting on a shared-nothing architecture using probabilistic splitting","authors":"D. DeWitt, J. Naughton, Donovan A. Schneider","doi":"10.1109/PDIS.1991.183115","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183115","url":null,"abstract":"The authors consider the problem of external sorting in a shared-nothing multiprocessor. A critical step in the algorithms the authors consider is to determine the range of sort keys to be handled by each processor. They consider two techniques for determining these ranges of sort keys: exact splitting, using a parallel version of the algorithm proposed by Iyer, Ricard, and Varman; and probabilistic splitting, which uses sampling to estimate quantiles. They present analytic results showing that probabilistic splitting performs better than exact splitting. Finally, the authors present experimental results from an implementation of sorting probabilistic splitting in the Gamma parallel database machine.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114123411","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183105
Yasuhiro Hirano, T. Satoh, Ushio Inoue, Katsumi Teranaka
This paper describes new load balancing algorithms for parallel database processing on shared memory multiprocessors. The goal of load balancing is to reduce overhead as well as load imbalance, but there is a tradeoff between them in ordinary algorithms. Unfortunately, optimum performance can hardly be obtained using ordinary algorithms because their performances depend on several factors such as database size, the number of processors and data distribution. The proposed algorithms solve these problems by varying the number of tasks allocated at a time ( which was fixed in ordinary algorithms ) according to the number of remaining tasks and the maximum and minimum processing times of a task. Performance evaluations show that the proposed algorithms achieve fair load balancing with lower overhead independent of the above factors.<>
{"title":"Load balancing algorithms for parallel database processing on shared memory multiprocessors","authors":"Yasuhiro Hirano, T. Satoh, Ushio Inoue, Katsumi Teranaka","doi":"10.1109/PDIS.1991.183105","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183105","url":null,"abstract":"This paper describes new load balancing algorithms for parallel database processing on shared memory multiprocessors. The goal of load balancing is to reduce overhead as well as load imbalance, but there is a tradeoff between them in ordinary algorithms. Unfortunately, optimum performance can hardly be obtained using ordinary algorithms because their performances depend on several factors such as database size, the number of processors and data distribution. The proposed algorithms solve these problems by varying the number of tasks allocated at a time ( which was fixed in ordinary algorithms ) according to the number of remaining tasks and the maximum and minimum processing times of a task. Performance evaluations show that the proposed algorithms achieve fair load balancing with lower overhead independent of the above factors.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114407837","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 : 1991-12-01DOI: 10.1109/PDIS.1991.183095
R. Marshall
Summary form only given. The authors describe a hardware and software architectural model for the characterization and analysis of genomes of select organisms. The data for research is obtained by semantically relating existing heterogeneous genetic databases. The authors have developed an architectural model based on associative memory structures at the hardware level and an object-oriented model at the software level. Simulation studies are under way to establish the appropriateness of a system based on a set of tagged associative processors operating in parallel which can be used for gene sequencing and analysis. They have identified primitive instructions on the machine which are consonant with biologists' map manipulation and usage requirements.<>
{"title":"A hardware and software architectural model for the manipulation of genetic databases","authors":"R. Marshall","doi":"10.1109/PDIS.1991.183095","DOIUrl":"https://doi.org/10.1109/PDIS.1991.183095","url":null,"abstract":"Summary form only given. The authors describe a hardware and software architectural model for the characterization and analysis of genomes of select organisms. The data for research is obtained by semantically relating existing heterogeneous genetic databases. The authors have developed an architectural model based on associative memory structures at the hardware level and an object-oriented model at the software level. Simulation studies are under way to establish the appropriateness of a system based on a set of tagged associative processors operating in parallel which can be used for gene sequencing and analysis. They have identified primitive instructions on the machine which are consonant with biologists' map manipulation and usage requirements.<<ETX>>","PeriodicalId":210800,"journal":{"name":"[1991] Proceedings of the First International Conference on Parallel and Distributed Information Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"1991-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133584365","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：