Norges Teknisk-naturvitenskapelige Universitet
نویسندگان
چکیده
This thesis presents a framework of a passively replicated transaction manager. By integrating transactions and replication, two well known fault tolerance techniques, the framework provides high availability for transactional systems and better support for non-deterministic execution for replicated systems. A prototype Java implementation of the framework, based on Jgroup/ARM and Jini, has been developed and performance tests have been executed. The results indicate that the response time for a simple credit-debit transaction heavily depends on the degree of replication for both servers and the transaction manager. E.g. a system with two replicas of the transaction manager and the servers quadruples the response time compared to the nonreplicated case. Thus, the performance penalty of replication should be weighed against the increased availability on a per application basis.
منابع مشابه
Norges Teknisk-naturvitenskapelige Universitet Fakultet for Informasjonsteknologi, Matematikk Og Elektroteknikk Hovedoppgave
.........................................................................................................................I PREFACE........................................................................................................................... II TABLE OF CONTENTS................................................................................................... III
متن کاملNORGES TEKNISK-NATURVITENSKAPELIGE UNIVERSITET A method of parametric solution of convolution equations
A variant of the method of moments is developed for parametric solution of convolution equations of the first kind. Two models — the gamma model and the shifted gamma model — are studied in details.
متن کاملNORGES TEKNISK-NATURVITENSKAPELIGE UNIVERSITET Implementing Approximate Bayesian Inference using Integrated Nested Laplace Approximation: a manual for the inla program
This manual describes the inla program, a new instrument which allows the user to easily perform approximate Bayesian inference using integrated nested Laplace approximation (INLA). We describe the set of models which can be solved by the inla program and provide a series of worked out examples illustrating its usage in details. Appendix A contains a reference manual for the inla program. This ...
متن کاملNORGES TEKNISK - NATURVITENSKAPELIGE UNIVERSITET Bayesian computing with INLA : new features by Thiago
The INLA approach for approximate Bayesian inference for latent Gaussian models has been shown to give fast and accurate estimates of posterior marginals and also to be a valuable tool in practice via the R-package R-INLA. In this paper we formalize new developments in the R-INLA package and show how these features greatly extend the scope of models that can be analyzed by this interface. We al...
متن کاملNorges Teknisk-naturvitenskapelige Universitet Cost Efficient Lie Group Integrators in the Rkmk Class Cost Efficient Lie Group Integrators in the Rkmk Class
In this work a systematic procedure is implemented in order to minimise the computational cost of the Runge–Kutta–Munthe-Kaas (RKMK) class of Lie-group solvers. The process consists of the application of a linear transformation to the stages of the method and the analysis of a graded free Lie algebra to reduce the number of commutators involved. We consider here RKMK integration methods up to o...
متن کاملNORGES TEKNISK-NATURVITENSKAPELIGE UNIVERSITET Parameter Estimation in High Dimensional Gaussian Distributions
In order to compute the log-likelihood for high dimensional Gaussian models, it is necessary to compute the determinant of the large, sparse, symmetric positive definite precision matrix. Traditional methods for evaluating the log-likelihood, which are typically based on Choleksy factorisations, are not feasible for very large models due to the massive memory requirements. We present a novel ap...
متن کامل