The authors present the case for allowing independent updates on replicated databases. In autonomous, heterogeneous, or large scale systems, using two-phase commit for updates may be infeasible. Instead, the authors propose that a site may perform updates independently. Sites that are available can receive these updates immediately. But sites that are unavailable, or otherwise do not participate in the update transaction receive these updates later through propagation, rather than preventing the execution of the update transaction until sufficient sites can participate. Two or more sites come to agreement using a reconciliation procedure that uses reception vectors to determine how much of the history log should be transferred from one site to another. They also consider what events can initiate a reconciliation procedure.<>
{"title":"The case for independent updates","authors":"S. Ceri, M. Houtsma, A. M. Keller, P. Samarati","doi":"10.1109/MRD.1992.242626","DOIUrl":"https://doi.org/10.1109/MRD.1992.242626","url":null,"abstract":"The authors present the case for allowing independent updates on replicated databases. In autonomous, heterogeneous, or large scale systems, using two-phase commit for updates may be infeasible. Instead, the authors propose that a site may perform updates independently. Sites that are available can receive these updates immediately. But sites that are unavailable, or otherwise do not participate in the update transaction receive these updates later through propagation, rather than preventing the execution of the update transaction until sufficient sites can participate. Two or more sites come to agreement using a reconciliation procedure that uses reception vectors to determine how much of the history log should be transferred from one site to another. They also consider what events can initiate a reconciliation procedure.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127217368","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}
B. Liskov, Sanjay Ghemawat, R. Gruber, Paul Johnson, L. Shrira
Harp is a replicated Unix file system accessible via the VFS interface. It provides highly available and reliable storage for files and guarantees that file operations are executed atomically in spite of concurrency and failures. Replication enables Harp to safely trade disk accesses for network communication and thus to provide good performance both during normal operation and during recovery. The authors focus on the techniques Harp uses to achieve efficient recovery.<>
{"title":"Efficient recovery in Harp (replicated Unix file system)","authors":"B. Liskov, Sanjay Ghemawat, R. Gruber, Paul Johnson, L. Shrira","doi":"10.1109/MRD.1992.242606","DOIUrl":"https://doi.org/10.1109/MRD.1992.242606","url":null,"abstract":"Harp is a replicated Unix file system accessible via the VFS interface. It provides highly available and reliable storage for files and guarantees that file operations are executed atomically in spite of concurrency and failures. Replication enables Harp to safely trade disk accesses for network communication and thus to provide good performance both during normal operation and during recovery. The authors focus on the techniques Harp uses to achieve efficient recovery.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126921624","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}
Replicated services can be implemented as process groups. Member processes use group communication protocols to communicate amongst themselves and group membership protocols to determine what processes are in the group. These protocols can provide various levels of consistency between members. The author investigates weak consistency protocols that guarantee that messages are delivered to all members, but do not guarantee when. He reports on a new family of communication protocols, an associated group membership mechanism, and current progress in evaluating their efficiency and utility for real applications.<>
{"title":"Weak consistency group communication for wide-area systems","authors":"Richard A. Golding","doi":"10.1109/MRD.1992.242627","DOIUrl":"https://doi.org/10.1109/MRD.1992.242627","url":null,"abstract":"Replicated services can be implemented as process groups. Member processes use group communication protocols to communicate amongst themselves and group membership protocols to determine what processes are in the group. These protocols can provide various levels of consistency between members. The author investigates weak consistency protocols that guarantee that messages are delivered to all members, but do not guarantee when. He reports on a new family of communication protocols, an associated group membership mechanism, and current progress in evaluating their efficiency and utility for real applications.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"373 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128846753","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}
Mirrored disk storage is an accepted technique to enhance fault-tolerance of data through complete replication. Recent research suggested that in addition to higher reliability, mirrored disks can offer better I/O performance by either allowing parallel reads or by reducing the seek time in cases where shortest seek distance algorithms can be used. Accurate analysis of mirrored disk systems shows that there is a significant correlation between the distributions of the disk heads positions, which makes them both interdependent and non-uniform. Furthermore, after each write all disk heads move to the same position to form what the authors call a 'bundle'. Such phenomenon may seriously deteriorate system performance. Subsequent reads gradually break up the bundles. Finally, under low rates of request arrivals, the authors provide analytical expressions for optimal anticipation points for two and three mirrored disks under general read/write combinations. Their work makes the following contributions: (1) new heuristics for treating the 'bundling phenomenon' adaptively, for applications with unknown read/write ratios; (2) a new technique that improves the expected seek time and workload balancing by using 'shifted mirroring'.<>
{"title":"I/O performance of fully-replicated disk systems","authors":"D. Rotem, Gerald A. Schloss","doi":"10.1109/MRD.1992.242614","DOIUrl":"https://doi.org/10.1109/MRD.1992.242614","url":null,"abstract":"Mirrored disk storage is an accepted technique to enhance fault-tolerance of data through complete replication. Recent research suggested that in addition to higher reliability, mirrored disks can offer better I/O performance by either allowing parallel reads or by reducing the seek time in cases where shortest seek distance algorithms can be used. Accurate analysis of mirrored disk systems shows that there is a significant correlation between the distributions of the disk heads positions, which makes them both interdependent and non-uniform. Furthermore, after each write all disk heads move to the same position to form what the authors call a 'bundle'. Such phenomenon may seriously deteriorate system performance. Subsequent reads gradually break up the bundles. Finally, under low rates of request arrivals, the authors provide analytical expressions for optimal anticipation points for two and three mirrored disks under general read/write combinations. Their work makes the following contributions: (1) new heuristics for treating the 'bundling phenomenon' adaptively, for applications with unknown read/write ratios; (2) a new technique that improves the expected seek time and workload balancing by using 'shifted mirroring'.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"176 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121081238","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The authors argue that, if they are going to assume low-sharing workloads, then it is possible to build more efficient implementations that provide the same (or stronger) semantics by making fullest use of the assumption. Accordingly, they take an extreme point on the spreading/searching tradeoff: the searcher is made responsible for all the work. Thus, spreading can be made asynchronous and hence very fast from the caller's viewpoint. They employ an important optimization to eliminate most of the searching work when there is little sharing. Their technique carries other advantages as well: simple and flexible recovery algorithms, and suitability for mobile computing.<>
{"title":"An interface to support lazy replicated file service","authors":"D. Duchamp, Carl D. Tait","doi":"10.1109/MRD.1992.242629","DOIUrl":"https://doi.org/10.1109/MRD.1992.242629","url":null,"abstract":"The authors argue that, if they are going to assume low-sharing workloads, then it is possible to build more efficient implementations that provide the same (or stronger) semantics by making fullest use of the assumption. Accordingly, they take an extreme point on the spreading/searching tradeoff: the searcher is made responsible for all the work. Thus, spreading can be made asynchronous and hence very fast from the caller's viewpoint. They employ an important optimization to eliminate most of the searching work when there is little sharing. Their technique carries other advantages as well: simple and flexible recovery algorithms, and suitability for mobile computing.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125268170","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The authors discuss extensions to the grid protocol to improve the fault-tolerance of write operations by using the notions of structured read and write grid quorums. As is the case in the standard quorum protocols, the increased fault-tolerance for write operations is at the increased cost of executing read operations. In order to let users continue using the analogues of the read-one-write-all protocol in the context of a logical structure, they develop reconfiguration protocols for dynamically adapting to failures and recovery. This results in the following dichotomy. Users accesses are through the simple analogues of the read-one-write-all protocol with respect to a logical structure and therefore have low communication cost for read operations. On the other hand, the reconfiguration protocol uses the notion of quorums in the context of a logical structure to ensure high data availability. A similar approach can be applied to the tree protocol.<>
{"title":"Dynamic logical structures: a position statement for managing replicated data","authors":"D. Agrawal, A. E. Abbadi","doi":"10.1109/MRD.1992.242624","DOIUrl":"https://doi.org/10.1109/MRD.1992.242624","url":null,"abstract":"The authors discuss extensions to the grid protocol to improve the fault-tolerance of write operations by using the notions of structured read and write grid quorums. As is the case in the standard quorum protocols, the increased fault-tolerance for write operations is at the increased cost of executing read operations. In order to let users continue using the analogues of the read-one-write-all protocol in the context of a logical structure, they develop reconfiguration protocols for dynamically adapting to failures and recovery. This results in the following dichotomy. Users accesses are through the simple analogues of the read-one-write-all protocol with respect to a logical structure and therefore have low communication cost for read operations. On the other hand, the reconfiguration protocol uses the notion of quorums in the context of a logical structure to ensure high data availability. A similar approach can be applied to the tree protocol.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134113003","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}
Andy Hisgen, A. Birrell, Charles Jerian, T. Mann, G. Swart
Understanding the workload is crucial for the success of a replicated file system. The system must continue to provide service in the presence of excess load, otherwise, its availability will be compromised. This paper presents some shortcomings of the Echo replicated file system in handling excess load. The authors discuss the consequences of excess load on primary secondary design, on their communication software, on election algorithm, on memory usage in their file servers, and on their distributed client caching algorithm. They speculate on possible solutions.<>
{"title":"Some consequences of excess load on the Echo replicated file system","authors":"Andy Hisgen, A. Birrell, Charles Jerian, T. Mann, G. Swart","doi":"10.1109/MRD.1992.242609","DOIUrl":"https://doi.org/10.1109/MRD.1992.242609","url":null,"abstract":"Understanding the workload is crucial for the success of a replicated file system. The system must continue to provide service in the presence of excess load, otherwise, its availability will be compromised. This paper presents some shortcomings of the Echo replicated file system in handling excess load. The authors discuss the consequences of excess load on primary secondary design, on their communication software, on election algorithm, on memory usage in their file servers, and on their distributed client caching algorithm. They speculate on possible solutions.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115107727","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}
Mobility of users and services and its impact on data replication and migration will be one of the main technical problems to be resolved. In fact, the authors view mobility as probably the most challenging new issue facing distributed systems of the future. The distinct new element which mobility brings into the issues of data replication is system's uncertainty about its own state (meaning primarily location). Thus the cost of communication between two mobile users is dependent not only on the distance between them but also on the cost of search necessary to determine the exact value of location. The authors address some of the questions, by proposing an adaptive replication scheme for mobile data.<>
{"title":"Replication and mobility","authors":"B. R. Badrinath, T. Imielinski","doi":"10.1109/MRD.1992.242628","DOIUrl":"https://doi.org/10.1109/MRD.1992.242628","url":null,"abstract":"Mobility of users and services and its impact on data replication and migration will be one of the main technical problems to be resolved. In fact, the authors view mobility as probably the most challenging new issue facing distributed systems of the future. The distinct new element which mobility brings into the issues of data replication is system's uncertainty about its own state (meaning primarily location). Thus the cost of communication between two mobile users is dependent not only on the distance between them but also on the cost of search necessary to determine the exact value of location. The authors address some of the questions, by proposing an adaptive replication scheme for mobile data.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126563866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The authors investigate the design of a replica management system (RMS) which allows a programmer to specify the quality of service required for individual replicated objects in terms of availability and performance. From the quality of service specification, information about the replication protocol to be used, and data about the characteristics of the underlying distributed system, the RMS computes an initial placement and replication level for the object. As machines and communications systems are detected to have failed or recovered, the RMS can be reinvoked to compute an updated mapping of replicas which preserves the desired quality of service. Preliminary simulation of the authors RMS shows that its placement algorithm gives consistent improvements in the availability of a replicated service than simply placing replicas on nodes at random, as is done in most distributed systems.<>
{"title":"Computing replica placement in distributed systems","authors":"D. McCue, M. Little","doi":"10.1109/MRD.1992.242617","DOIUrl":"https://doi.org/10.1109/MRD.1992.242617","url":null,"abstract":"The authors investigate the design of a replica management system (RMS) which allows a programmer to specify the quality of service required for individual replicated objects in terms of availability and performance. From the quality of service specification, information about the replication protocol to be used, and data about the characteristics of the underlying distributed system, the RMS computes an initial placement and replication level for the object. As machines and communications systems are detected to have failed or recovered, the RMS can be reinvoked to compute an updated mapping of replicas which preserves the desired quality of service. Preliminary simulation of the authors RMS shows that its placement algorithm gives consistent improvements in the availability of a replicated service than simply placing replicas on nodes at random, as is done in most distributed systems.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127068298","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The authors describe their experience in attempting to modularize Consul, a fault-tolerant system for managing replicated data. Their experience is that, while modularity is needed to simplify the design and implementation of such systems, protocol dependencies, both direct and indirect, impact the way in which protocols are designed and implemented. They also identify certain operating system features that simplify such a design. Based on this experience, work is in progress to develop a new model for fault-tolerant protocols that will facilitate modularization.<>
{"title":"Protocol modularity in systems for managing replicated data","authors":"Shivakant Mishra, L. Peterson, R. Schlichting","doi":"10.1109/MRD.1992.242612","DOIUrl":"https://doi.org/10.1109/MRD.1992.242612","url":null,"abstract":"The authors describe their experience in attempting to modularize Consul, a fault-tolerant system for managing replicated data. Their experience is that, while modularity is needed to simplify the design and implementation of such systems, protocol dependencies, both direct and indirect, impact the way in which protocols are designed and implemented. They also identify certain operating system features that simplify such a design. Based on this experience, work is in progress to develop a new model for fault-tolerant protocols that will facilitate modularization.<<ETX>>","PeriodicalId":314844,"journal":{"name":"[1992 Proceedings] Second Workshop on the Management of Replicated Data","volume":"368 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1992-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116622013","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：