• Journal of Korea Multimedia Society
• /
• v.7 no.8
• /
• pp.1097-1106
• /
• 2004

The typical web cluster architecture consists of replicated real servers and a virtual server that routes client requests among the real servers. In this paper, we proposed an efficient dynamic load distribution method with load prediction for the web cluster systems. The virtual server transmit status request message to real servers in other to get load states. However the load states dose not accurate during load distribution, thus the virtual server predict the load status of real servers and assign a request of the client to the minimum loaded real server. The proposed distribution methods works not related to partial breakdown of real servers, thus the system works with high availability. We also show that the proposed distribution method preserve scalable property and improve the throughput through a set of simulations.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.04a
• /
• pp.776-779
• /
• 2009

P2P(Peer-to-Peer) 기반에 분산 작업 할당의 경우 정해진 수의 모든 피어에 작업을 분배하는 방식으로 처리하기 때문에 한 피어에 작업량이 증가하는 경우나 피어에 문제가 발생하는 경우 이를 사전에 예측하지 못해 시스템에 문제가 발생하게 된다. 또한 P2P 형태의 특성 상 피어의 그룹 참여는 유동적이며 그룹에 참여하는 피어의 사양도 다양하게 존재하게 된다. 이와 같은 상황에서는 작업을 할당하더라도 작업 진행이 이루어지지 않을 확률이 높아지고 전반적으로 시스템의 성능을 저하시키는 문제점을 가지게 된다. 본 논문에서는 이러한 문제점을 해결하기 위해서 인공지능 알고리즘의 하나인 유전자 알고리즘(Genetic Algorithms: GA)를 활용하여 피어의 상태를 사전에 예측하여 작업 분배 시 동적으로 작업할 피어를 선택하여 처리하는 P2P기반 동적 작업할당 관리자를 설계하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2003.05a
• /
• pp.109-112
• /
• 2003

고성능의 웹 서비스를 지원하기 위하여 LVS 시스템을 구축하는 경우가 증가하고 있다. LVS 시스템에는 각 RealServer에 부하를 적절히 분산하여 시스템의 구성을 유지하는 부하 분배작업이 매우 중요하다. 본 논문에서는 SNMP 메시지를 사용하여 분배서버가 RealServer의 부하 상태를 획득할 수 있도록 구성하였다. 또한, 분배서버가 메모리 사용량, 부하량의 정보를 각 RealServer로부터 주기적으로 수집하고 이 정보를 바탕으로 동적 가중치를 계산한 뒤 가중치 갱신을 위한 주기 사이의 실시간 접속자 수를 부가하여 RealServer를 선정하는 동적 부하 분산 방법을 제안한다. 제안된 방법은 간단한 구성의 시스템에서 실험되었으며 잘 동작하는 것으로 확인되었다.

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.11
• /
• pp.831-835
• /
• 2009

Volunteer computing is a computing paradigm in which the operations of a large-scale application are processed using idle resources of Internet-connected computers. The redundant-work distribution techniques used in the volunteer computing environment cause some problems like delay of the work completion time or inefficiency of the work execution due to continuous requests of the work redistribution because the techniques do not consider the characteristics of works and dynamic resources. To cope with such problems this paper suggests a P2P based regional matchmaking technique which can redistribute works in consideration of the characteristics of the works and the participant resources. The techniques consist of the profiles for works/resources, the module for monitoring the available resources spaces, the protocols for the matching request/reply, the module for match-ratio calculation, and so on.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2017.04a
• /
• pp.26-29
• /
• 2017

최근 중앙 집중화된 대규모 클라우드 시스템의 증가로 인해 가상화 환경에서 수행되는 성능 최적화 작업에 대한 많은 연구가 진행되고 있다. 그러나 기존 연구에서는 자원 분배의 공정성을 위해 가상머신 단위로 컴퓨팅 자원을 격리한 정책 내에서 이루어지고 있어 유연한 자원관리에 한계를 가지고 있다. 본 연구에서는 워크로드의 특징을 기반으로 과학적 연산을 수행하는 가상머신과 일반적인 서비스를 수행하는 가상머신을 분류하여 성능 최적화 작업을 수행하는 동적 자원 관리자를 제안한다. 실험을 통하여 제안하는 동적 자원 관리자가 KVM 기본 스케줄링에 비해 49%의 성능 향상을 보였다.

• Journal of the Korea Society for Simulation
• /
• v.18 no.2
• /
• pp.29-38
• /
• 2009

In the computing environment with heterogeneous resources, a job scheduling model is necessary for effective resource utilization and high-speed data processing. And, the job scheduling model has to cope with a dynamic change in the condition of resources. There have been lots of researches on resource estimation methods and heuristic algorithms about how to distribute and allocate jobs to heterogeneous resources. But, existing researches have a weakness for system compatibility and scalability because they do not support the standard language. Also, they are impossible to process jobs effectively and deal with a variety of computing situations in which the condition of resources is dynamically changed in real-time. In order to solve the problems of existing researches, this paper proposes a semantic computing-based dynamic job scheduling model that defines various knowledge-based rules for job scheduling methods adaptable to changes in resource condition and allocate a job to the best suited resource through inference. This paper also constructs a resource ontology to manage information about heterogeneous resources without difficulty as using the OWL, the standard ontology language established by W3C. Experimental results shows that the proposed scheduling model outperforms existing scheduling models, in terms of throughput, job loss, and turn around time.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.20 no.3
• /
• pp.83-88
• /
• 2020

With developing vehicle and communication technologies, cars can communicate with road-side infrastructures and among other cars. As autonomous driving cars have been developed, the cars are equipped with many sensors and powerful processing units. There are many studies related to provide cloud services to users by using available resources of connected cars. In this paper, we proposed a dynamic task size decision scheme that considers communication environment between a vehicle and a base station as well as available resources while allocating a proper task to each vehicle. Simulation results based on the proposed model show that a vehicle can complete its allocated task when we considers available resources and communication environments.

• The KIPS Transactions:PartA
• /
• v.14A no.6
• /
• pp.357-362
• /
• 2007

In distributed systems like cluster systems, in order to get more performance improvement, the initial task placement system precisely estimates and correctly assigns the resource requirement by the process. The resource-based initial job placement scheme needs the prediction of resource usage of a task in order to fit it to the most suitable hosts. However, the wrong prediction of resource usage causes serious performance degradation in dynamic load balancing systems. Therefore, in this paper, to resolve the problem due to the wrong prediction, we propose a new load metric. By the new load metric, the resource-based initial job placement scheme can work without priori knowledge about the type of process. Simulation results show that the dynamic load balancing system using the proposed approach achieves shorter execution times than the conventional approaches.

• Journal of KIISE:Computing Practices and Letters
• /
• v.8 no.2
• /
• pp.183-194
• /
• 2002

The clustering environment with heterogeneous workstations provides the cost effectiveness and usability for executing applications in parallel. Load balancing is considered a necessary feature for a cluster of heterogeneous workstations to minimize the turnaround time. Previously, static load balancing that assigns a predetermined weight for the processing capability of each workstation, or dynamic approaches which execute a benchmark program to get relative processing capability of each workstation were proposed. The execution of the benchmark program, which has nothing to do with the application being executed, consumes the computation time and the overall turnaround time is delayed. In this paper, we present efficient methods for task distribution and task migration, based on the relative load index. We designed and implemented a load balancing system for the clustering environment with heterogeneous workstations. Turnaround times of our methods and the round-robin approach, as well as the load balancing method using a benchmark program, were compared. The experimental results show that our methods outperform all the other methods that we compared.

• Journal of Internet Computing and Services
• /
• v.2 no.3
• /
• pp.81-92
• /
• 2001

The clustering environment with heterogeneous workstations provides the cost effectiveness and usability for executing applications in parallel. The load balancing is considered as a necessary feature for the clustering of heterogeneous workstations to minimize the turnaround time. Since each workstation may have different users, groups. requests for different tasks, and different processing power, the capability of each processing unit is relative to the others' unit in the clustering environment Previous works is a static approach which assign a predetermined weight for the processing capability of each workstation or a dynamic approach which executes a benchmark program to get relative processing capability of each workstation. The execution of the benchmark program, which has nothing to do with the application being executed, consumes the computation time and the overall turnaround time is delayed. In this paper, we present an efficient task distribution method and implementation of load balancing system for the clustering environment with heterogeneous workstations. Turnaround time of the methods presented in this paper is compared with the method without load balancing as well as with the method load balancing with performance evaluation program. The experimental results show that our methods outperform all the other methods that we compared.