CLUSTERED SIMULATION EXPERIMENTER: A tool for concurrent simulation execution on loosely coupled workstations
نویسندگان
چکیده
The "Clustered Simulation Experimenter" (CSX) project intends to make massive concurrent experiments without having to be concerned in locating computational resources to carry it out. This paper presents a tool whose principal purpose is to use a lot of idle and heterogeneous workstations at university laboratories and research centers to run concurrent simulations. The approach consists of adding and eliminating machines, with different architectures and operating systems, to a heterogeneous and dynamic cluster that permits to execute a distributed application. The communication among the machines and the co-ordination of the distributed application processes uses a typical network infrastructure using PVM v3.4. (Parallel Virtual Machine). An asynchronous design of the communications lets to execute sequential simulation models as fast as running it as normal simulation programs. Simulations executed using CSX are classic discrete-event oriented models that, basically, will be run using replication techniques, and monitored for extracting statistic information of the output variables of the model, letting CSX to improve response times through possibilities of an improved statistic output analysis. The first effort has been to be able to monitor the output variables of the models without causing an excessive "overhead". We present here how has been applied to SMPL, but the purpose is to use it in simulators such as CSIM, SIM++, QNAP and DEVS*. Due to the fact that the distributed execution environment is highly variable and is shared with other users, it is necessary to apply dynamic load balancing techniques.
منابع مشابه
Investigation on Reliability Estimation of Loosely Coupled Software as a Service Execution Using Clustered and Non-Clustered Web Server
Evaluating the reliability of loosely coupled Software as a Service through the paradigm of a cluster-based and non-cluster-based web server is considered to be an important attribute for the service delivery and execution. We proposed a novel method for measuring the reliability of Software as a Service execution through load testing. The fault count of the model against the stresses of users ...
متن کاملDIVIDE: Distributed Visual Display of the Execution of Asynchronous, Distributed Algorithms on Loosely-Coupled Parallel Processors
The issue of monitoring the execution of asynchronous, distributed algorithms on loosely-coupled parallel processor systems, is important for the purposes of (i) detecting inconsistencies and aws in the algorithm, (ii) obtaining important performance parameters for the algorithm, and (iii) developing a conceptual understanding of the algorithm's behavior, for given input stimulus, through visua...
متن کاملNimrod: A Tool for Performing Parametised Simulations using Distributed Workstations
This paper discusses Nimrod, a tool for performing parametised simulations over networks of loosely coupled workstations. Using Nimrod the user interactively generates a parametised experiment. Nimrod then controls the distribution of jobs to machines and the collection of results. A simple graphical user interface which is built for each application allows the user to view the simulation in te...
متن کاملA Tool for Performing Parametised Simulations using Distributed Workstations
This paper discusses Nimrod, a tool for performing parametised simulations over networks of loosely coupled workstations. Using Nimrod the user interactively generates a parametised experiment. Nimrod then controls the distribution of jobs to machines and the collection of results. A simple graphical user interface which is built for each application allows the user to view the simulation in te...
متن کاملNimrod: A Tool for Performing Parameterised Simulations Using Distributed Workstations
This paper discusses Nimrod, a tool for performing parametised simulations over networks of loosely coupled workstations. Using Nimrod the user interactively generates a parametised experiment. Nimrod then controls the distribution of jobs to machines and the collection of results. A simple graphical user interface which is built for each application allows the user to view the simulation in te...
متن کامل