For efficient construction of the distributeddatabase from media failure, a transaction-consistent checkpointing algorithm is proposed for heterogeneous distributeddatabasesystems. For full design autonomy and incr...
详细信息
For efficient construction of the distributeddatabase from media failure, a transaction-consistent checkpointing algorithm is proposed for heterogeneous distributeddatabasesystems. For full design autonomy and increased availability on the heterogeneous distributeddatabasesystems. the proposed algorithm never enforces termination of normal operations of transactions and changes of checkpointing algorithms in local databasesystems. The global checkpoints generated by the algorithm can be used to reconstruct the previous consistent states of a database efficiently.< >
Presents a feasibility study of two combined techniques for software fault tolerance in distributedsystems. A probabilistic model for each technique is presented that represents a technique's ability to recover f...
详细信息
Presents a feasibility study of two combined techniques for software fault tolerance in distributedsystems. A probabilistic model for each technique is presented that represents a technique's ability to recover from software and hardware faults that occur with arbitrary frequency. The goal is not to accurately predict the behavior of systems.that employ these techniques, but rather to identify the strong and weak points of each approach under varying circumstances of dependability in both the system and the fault-tolerance components. An example study shows how these models can be used to understand the influence of component failures on the reliability of systems.using these two approaches.< >
Although many solutions to the problem of representing spatial data have been proposed, most are not practical. They cannot be guaranteed to perform well with large and arbitrarily distributed data collections. They m...
详细信息
Although many solutions to the problem of representing spatial data have been proposed, most are not practical. They cannot be guaranteed to perform well with large and arbitrarily distributed data collections. They may not be easily integrated with concurrency and recovery software already written for databasesystems. However, two proposed structures do have some guarantees, similar to those which have made the B/sup +/-tree so successful for one-dimensional data. These analytic guarantees include worst case space utilization in data and index pages, a minimal fan-out and exact match search time bounded by the height of the tree. The two methods are the holey brick tree and bit interleaving using a B/sup +/-tree. Both can also be integrated with concurrency and recovery systems.in the same way that B/sup +/-trees are.< >
The proposed coordinator log transaction execution protocol centralizes logging on a per-transaction basis and exploits piggybacking to provide the semantics of a distributed atomic commit without the associated costs...
详细信息
The proposed coordinator log transaction execution protocol centralizes logging on a per-transaction basis and exploits piggybacking to provide the semantics of a distributed atomic commit without the associated costs. This protocol eliminates two rounds of messages (one phase) from the presumed commit protocol and dramatically reduces the number of log forces needed for distributed atomic commit. The authors compare the coordinator log transaction execution protocol with existing protocols, describe when it is desirable, and discuss how it affects the write-ahead log protocol and the database crash recovery algorithm.
distributed synchronization for data sharing is discussed, and the design of a distributed lock manager for the Camelot transaction facility is presented. The lock manager is a component of a proposed implementation o...
详细信息
distributed synchronization for data sharing is discussed, and the design of a distributed lock manager for the Camelot transaction facility is presented. The lock manager is a component of a proposed implementation of data sharing in the Camelot environment. A number of experiments that demonstrate the correct operation of the lock manager are reported and its performance is described. The performance metrics indicate that distributed lock management should not reduce the feasibility of data sharing in this environment. The similarity between the caching and synchronization strategies appropriate for locks and data suggests that protocols developed for distributed locks will be applicable to data sharing.
A recursive algorithm for computing a lower bound on the all-terminal reliability of an n-dimensional hypercube is presented. The recursive step decomposes an n-dimensional hypercube into lower dimension hypercubes th...
详细信息
A recursive algorithm for computing a lower bound on the all-terminal reliability of an n-dimensional hypercube is presented. The recursive step decomposes an n-dimensional hypercube into lower dimension hypercubes that are linked together. As an illustration of the effectiveness and power of this method, a lower bound is computed on the all-terminal reliability of the 16-dimensional hypercube (Connection Machine architecture) whose links number 219. The notation and assumptions are defined, and background information on bounding the reliability polynomial is provided. Methods for tightening these bounds for the analysis of the hypercube architecture are discussed.
The following topics are dealt with: distributed operating systems.local area networks;network fault tolerance;hypercubes;distributeddatabases;real-time systems.replicated programs;computer architectures;and voting. ...
详细信息
The following topics are dealt with: distributed operating systems.local area networks;network fault tolerance;hypercubes;distributeddatabases;real-time systems.replicated programs;computer architectures;and voting. Abstracts of individual papers can be found under the relevant classification codes in this or other issues.
Replicated execution of distributed programs, which provides a means of masking hardware (processor) failures in a distributed system, is discussed. Application-level entities (processes, objects) are replicated to ex...
详细信息
Replicated execution of distributed programs, which provides a means of masking hardware (processor) failures in a distributed system, is discussed. Application-level entities (processes, objects) are replicated to execute on distinct processors. Such replica entities communicate by message passing. Nondeterminism within the replicas could cause messages to be processed in nonidentical order, producing a divergence of state. Possible sources of nondeterminism are identified, and a generic mechanism for ensuring that nonfaulty replicas process messages in identical order, thereby preventing state divergence among such replicate entities, is presented.
A knowledge-based approach for query processing during network partitioning is proposed. The approach uses available domain and summary knowledge to infer inaccessible data to answer a given query. A rule induction te...
详细信息
A knowledge-based approach for query processing during network partitioning is proposed. The approach uses available domain and summary knowledge to infer inaccessible data to answer a given query. A rule induction technique is used to extract correlated knowledge between attributes from the database contents. This knowledge is represented as rules for data inference. On the basis of a set of queries, simulation is used to evaluate the effectiveness of the proposed data inference technique for improving data availability under network partitioning. Object allocation has a significant impact on data availability. Allocating objects that increase remote redundancy and reduce local redundancy increases data availability during network partitioning. A prototype distributeddatabase system that uses the proposed inference technique with correlated knowledge from a ship database has been implemented. Experience indicates that the proposed inference technique can significantly improve the availability of a distributeddatabase during network partitioning.
The problem of voting is studied for both the exact and inexact cases. Optimal solutions based on explicit computation of condition probabilities are given. The most commonly used strategies, i.e., majority, median, a...
详细信息
The problem of voting is studied for both the exact and inexact cases. Optimal solutions based on explicit computation of condition probabilities are given. The most commonly used strategies, i.e., majority, median, and plurality are compared quantitatively. The results show that plurality voting is the most powerful of these techniques and is, in fact, optimal for a certain class of probability distributions. An efficient method of implementing a generalized plurality voter when nonfaulty processes can produce differing answers is also given.
暂无评论