3D Analytical Modelling and Iterative Solution for High Performance Computing Clusters
ABSTARCT :
Mobile Cloud Computing enables the migration of services to the edge of Internet. Therefore, high performance computing clusters are widely deployed to improve computational capabilities of such environments. However, they are prone to failures and need analytical models to predict their behaviour in order to deliver desired quality-of-service and quality-of-experience to mobile users. This paper proposes a 3D analytical model and a problem-solving approach for sustainability evaluation of high-performance computing clusters. The proposed solution uses an iterative approach to obtain performance measurements to overcome the state space explosion problem. The availability modelling and evaluation of master and computing nodes are performed using a multi-repairman approach. The optimum number of repairmen is also obtained to get realistic results and reduce the overall cost. The proposed model is validated using discrete event simulation. The analytical approach is much faster and in good agreement with the simulations. The analysis focuses on mean queue length, throughput and mean response time outputs. The maximum differences between analytical and simulation results in the considered scenarios of up to a billion states are less than 1.149%, 3.82%, and 3.76%, respectively. These differences are well within the 5% of confidence interval of the simulation and the proposed model.
EXISTING SYSTEM :
? Existing modelling and exact solution approaches such as the spectral expansion method is not able to handle such large networks.
? In a situation of the state space explosion problem is encountered in a spectral expansion method where Beowulf clusters are modelled and solved for various performance measures.
? HPC systems fundamentally provide access to large pools of data and computational resources through a variety of interfaces similar to the existing grid, HPC resource management and programming systems.
? The HPC service providers must strive to ensure good QoS by offering highly available services with dynamically scalable resources as stated in.
DISADVANTAGE :
? The state space explosion is a major problem for such modelling because it limits the size of such systems that can be evaluated.
? The proposed iterative solution approach has been shown to overcome the state space explosion problem for steady-state queuing systems while giving good approximate results.
? Therefore, it is of great importance to develop an analytical method and a solution approach to overcome these problems for such systems.
? The proposed model uses an iterative solution approach and a large number of nodes with failure of master and computing nodes can be handled without causing the state space explosion problem.
? However, large-scale clusters could not be considered due to the state space explosion problem.
PROPOSED SYSTEM :
• The proposed analytical model and the approximate solution approach provide flexibility to evaluate the QoS measurements for such systems.
• The performability results obtained from the analytical model are compared to the discrete event simulation (DES) results in order to show the accuracy and effectiveness of the proposed work.
• However, the proposed models and solution approaches used are more applicable to small and/or medium size systems rather than large-scale systems.
• The CPU times of the proposed solution approach and DES are also compared in order to show the efficacy of the proposed model and an approximate solution approach.
ADVANTAGE :
? High performance computing (HPC) has therefore emerged as an appealing MCC service, especially with the proliferation of big data in a mobile environment because it can be used to provide reduced latency in an MCC environment and thus results in high performance.
? In many practical systems, failures are expected and they can significantly affect the system performance.
? Therefore, when the number of nodes is significantly large, focusing purely on performance without taking into account the possibility of node failures would lead to a significant over-estimation in how an actual system would perform.
? Analytical performance and availability models have been developed for large-scale HPCCs in a MCC environment.
|