Application-specific Cloud Provisioning Model Using Job Profiles Analysis

Abstract

The rapid advent of computing technology enables scientists to expand their research infrastructure over supercomputer-on-demand using a resource leasing service on pay-per-use basis. This infrastructure service is called as Science cloud which provides uniform user interface to scientific experiments over large-scale heterogeneous resources. In spite of the strength of conveniences, it is difficult to manage the experiments to guarantee optimal performance of jobs since the execution environment is based on virtualization technology which has additional performance overheads.This paper proposes a cloud resource provisioning model using statistical analysis of job profiles for computational science. In this model, we utilize job profiles which are generated from application executions and identify features of application by applying statistical analysis. The analysis is performed using PCA and the most effective factors are inferred from profiles. The effective factors are used for picking a job profile being referred and then VM type is determined using the most effective job s property. An application is executed on VM of the chosen cluster and its performance result is incorporated into job profile set with purpose of evaluating profile's credit. Our simulation results show this model enables scientists to take advantages of cloud computing without performance degradation.

Resource Scheduling Techniques in Cloud Computing Environment : A Survey

“…There are many resource provisioning techniques both static and dynamic provisioning each having its own pros and cons. The provisioning techniques are used to improve QoS parameters [1,3,5,13], minimize cost for cloud user and maximize revenue for the Cloud Service Provider [17,25,30,32,38], improve response time [2], deliver services to cloud user even in presence of failures [11,14,15,31], improve performance [26,28,29,33,34,36] reduces SLA violation [27,37], efficiently uses cloud resources [4,6,9,10,16,19,20,22,24,40], reduces power consumption [7,21].…”

Value and Energy Aware Adaptive Resource Allocation of Soft Real-Time Jobs on Many-Core HPC Data Centers

“…These profiling results are used to facilitate lightweight run-time resource allocation as the compute intensive part is shifted to design-time. Such allocation approaches have been proven promising to design job-specific-clouds, where the clients (or customers) and their jobs to be submitted for execution are pre-defined, which can be realized from the historical data [11], [12]. However, these approaches do not perform run-time adaptive allocation by monitoring the availability of cores on different servers and execution status of the jobs.…”

“…Since jobs are profiled in advance, it is assumed that all the clients (or customers) and their jobs to be submitted for execution in the HPC data center are known in advance. This is true for several data centers as they serve only a fixed set of known customers and such information facilitates to design promising job-specific-clouds [11], [12]. The profiling results are used to perform efficient run-time allocation and adaptation.…”