• Title/Summary/Keyword: Grid Scheduling

Search Result 142, Processing Time 0.031 seconds

A Resource Information Model for High Performance GRID Environemnts (고성능 그리드 환경을 위한 자원정보모델에 관한 연구)

  • Kim Hie-Cheol;Lee Kang-Woo;Lee Yong-Doo;Cho Sae-Hong
    • Journal of Digital Contents Society
    • /
    • v.2 no.2
    • /
    • pp.167-178
    • /
    • 2001
  • For high performance Grid environments, an effective GIS(Grid Information System) should be adopted. In the design of GIS architecture, its grid resource information model provides a key basis. This paper presents our study on the exploration of a high performance grid resource information model. According the exploration, we identified the followings. The resource information model should clearly address the issues of relation descriptions as well as resource descriptions, issues related to scheduling support, the issue of decoupling the expression model of resource information from data repository models, and finally the issue of decoupling user-level resource descriptions from system-level resource descriptions. Based on the proposed conceptual organization of resource information models, analysis result for the existing resource information models are presented.

  • PDF

Application Contents Deploy System based on workflow process a unit based for collaboration in Grid Environment (그리드 환경에서 협업을 위한 워크플로우 프로세스 단위 기반의 애플리케이션 컨텐츠 배포 시스템)

  • Moon, Seok-Jae;Heo, Hyuk;Choi, Young-Keun
    • Journal of the Korea Institute of Information and Communication Engineering
    • /
    • v.12 no.1
    • /
    • pp.172-182
    • /
    • 2008
  • Application Contents of based on a workflow process unit in grid system is a practical work to solve the problem. To construct cooperation environment is very important because the contents are deployed and executed. And the most of cooperation is executed by workflow. Globus toolkit is the middleware for cooperation process in grid environment. However the middleware only provides basic services for constructing grid, not workflow that can provides cooperation, job scheduling, application contents management. Also Globus toolkit is suitable for huge gird community but small size pc based grid community. Therefore, this paper proposes the efficient application contents deployment system for small sizes cooperation. This system supports efficient cooperation environment with Application contents deployment and resource management provided in grid community. Also it presents information and association relation among application contents with association dictionary among application contents based on workflow process unit, and utilizes this as information for application contents deployment.

Collaborative Inference for Deep Neural Networks in Edge Environments

  • Meizhao Liu;Yingcheng Gu;Sen Dong;Liu Wei;Kai Liu;Yuting Yan;Yu Song;Huanyu Cheng;Lei Tang;Sheng Zhang
    • KSII Transactions on Internet and Information Systems (TIIS)
    • /
    • v.18 no.7
    • /
    • pp.1749-1773
    • /
    • 2024
  • Recent advances in deep neural networks (DNNs) have greatly improved the accuracy and universality of various intelligent applications, at the expense of increasing model size and computational demand. Since the resources of end devices are often too limited to deploy a complete DNN model, offloading DNN inference tasks to cloud servers is a common approach to meet this gap. However, due to the limited bandwidth of WAN and the long distance between end devices and cloud servers, this approach may lead to significant data transmission latency. Therefore, device-edge collaborative inference has emerged as a promising paradigm to accelerate the execution of DNN inference tasks where DNN models are partitioned to be sequentially executed in both end devices and edge servers. Nevertheless, collaborative inference in heterogeneous edge environments with multiple edge servers, end devices and DNN tasks has been overlooked in previous research. To fill this gap, we investigate the optimization problem of collaborative inference in a heterogeneous system and propose a scheme CIS, i.e., collaborative inference scheme, which jointly combines DNN partition, task offloading and scheduling to reduce the average weighted inference latency. CIS decomposes the problem into three parts to achieve the optimal average weighted inference latency. In addition, we build a prototype that implements CIS and conducts extensive experiments to demonstrate the scheme's effectiveness and efficiency. Experiments show that CIS reduces 29% to 71% on the average weighted inference latency compared to the other four existing schemes.

Scheduling and Load Balancing Methods of Multithread Parallel Linear Solver of Finite Element Structural Analysis (유한요소 구조해석 다중쓰레드 병렬 선형해법의 스케쥴링 및 부하 조절 기법 연구)

  • Kim, Min Ki;Kim, Seung Jo
    • Journal of the Korean Society for Aeronautical & Space Sciences
    • /
    • v.42 no.5
    • /
    • pp.361-367
    • /
    • 2014
  • In this paper, task scheduling and load balancing methods of multifrontal solution methods of finite element structural analysis in a modern multicore machine are introduced. Many structural analysis problems have generally irregular grid and many kinds of properties and materials. These irregularities and heterogeneities lead to bottleneck of parallelization and cause idle time to analysis. Therefore, task scheduling and load balancing are desired to reduce inefficiency. Several kinds of multithreaded parallelization methods are presented and comparison between static and dynamic task scheduling are shown. To reduce the idle time caused by irregular partitioned subdomains, computational load balancing methods, Balancing all tasks and minmax task pairing balancing, are invented. Theoretical and actual elapsed time are shown and the reason of their performance gap are discussed.

An Optimal Energy Storage Operation Scheduling Algorithm for a Smart Home Considering Life Cost of Energy Storage System

  • Yan, Luo;Baek, Min-Kyu;Park, Jong-Bae;Park, Yong-Gi;Roh, Jae Hyung
    • Journal of Electrical Engineering and Technology
    • /
    • v.12 no.4
    • /
    • pp.1369-1375
    • /
    • 2017
  • This paper presents an optimal operation scheduling algorithm for a smart home with energy storage system, electric vehicle and distributed generation. The proposed algorithm provides the optimal charge and discharge schedule of the EV and the ESS. In minimizing the electricity costs of the smart home, it considers not only the cost of energy purchase from the grid but also the life cost of batteries. The life costs of batteries are calculated based on the relation between the depth of discharge and life time of battery. As the life time of battery depends on the charge and discharge pattern, optimal charge and discharge schedule should consider the life cost of batteries especially when there is more than one battery with different technical characteristics. The proposed algorithm can also be used for optimal selection of size and type of battery for a smart home.

Unit Commitment of a Microgrid Considering Islanded Operation Scenarios (독립운전 시나리오를 고려한 마이크로그리드의 최적 발전기 기동정지 계획)

  • Lee, Si Young
    • The Transactions of The Korean Institute of Electrical Engineers
    • /
    • v.67 no.6
    • /
    • pp.708-714
    • /
    • 2018
  • Islanded operation of a microgrid can ensure the reliable operation of the system when a large accident occurs in the main grid. However, because the generation capability of a microgrid is typically limited, a microgrid operator should take islanded operation risk into account in scheduling its generation resources. To address this problem, in this paper we have proposed two unit commitment formulations based on the islanding scenario that reflect the expected and worst-case values of the islanded operation risk. An optimal resource scheduling strategy is obtained for the microgrid operator by solving these optimization problem, and the effectiveness of the proposed method is investigated by numerical simulations.

A Method on the Realization of QoS Guarantee in the Grid Network (그리드 네트워크에서의 QoS 보장방법 구현)

  • Kim, Jung-Yun;Na, Won-Shin;Ryoo, In-Tae
    • Journal of Digital Contents Society
    • /
    • v.10 no.1
    • /
    • pp.169-175
    • /
    • 2009
  • Grid computing is an application to obtain the most efficient performance from computing resources in terms of cost and convenience. It is also considered as a good method to solve a problem that cannot be settled by conventional computing technologies such as clustering or is requiring supercomputing capability due to its complex and long-running task. In order to run grid computing effectively, it needs to connect high-performance computing resources in real-time which are distributed geographically. Answering to the needs of this grid application, researchers in several universities with Argonne National Laboratory in the USA (ANL) as the main axis have developed Globus. It is noticed, however, that the quality of service (QoS) is not guaranteed when certain jobs are exchanged through networks in the context of Globus. To tackle with this problem, the ANL has invented Globus Architecture for Reservation and Allocation (GARA). The researchers of this paper constructed a testbed for evaluating the ability to reserve resource in the GARA system and implemented the GARA code for it. We analyzed the applied results and discussed future research plans.

  • PDF

A Job Scheduling Mechanism Using Multi-queue In Grid Computing Environments (그리드 컴퓨팅 환경에서 다중 큐를 이용한 작업 스케줄링 기법)

  • Kang, Chang-Hoon;Park, Kie-Jin;Kim, Sung-Soo
    • Proceedings of the Korean Information Science Society Conference
    • /
    • 2005.11a
    • /
    • pp.562-564
    • /
    • 2005
  • 최근 지역적으로 분산되어 있는 이질적인 고성능 컴퓨팅 자원을 하나로 묶어 거대한 시스템을 구성하는 그리드 컴퓨팅(Grid Computing)에 대한 연구가 활발하게 이루어짐에 따라, 본 논문에서는 그리드 시스템에서 클라이언트 작업들을 특성에 따라 분류하여 우선순위가 높은 작업은 예약이 가능한 큐로 분배하고 우선순위가 낮은 작업은 백필이 가능한 큐로 할당하는 다중 큐 스케줄링 기법을 제안하였고 다양한 실험을 통하여 제안된 기법들의 성능을 평가하였다. 그 결과 그리드 컴퓨팅 시스템의 이용률(Utilization)이 높아지고, 작업 지연시간(Slowdown)이 줄어드는 것을 확인하였다.

  • PDF

A Job Scheduling S초듣 of Grid Computing using Hybrid Multi-queue (혼합 다중 큐를 사용한 그리드 컴퓨팅의 작업 스케줄링 기법)

  • Kang Chang-Hoon;Choi Chang-Yeol;Park Kie-Jin;Kim Sung-Soo
    • Annual Conference of KIPS
    • /
    • 2006.05a
    • /
    • pp.757-760
    • /
    • 2006
  • 고성능 컴퓨팅 자원(Computing Resources)을 네트워크를 통하여 하나로 연동하여 단일 컴퓨팅 작업을 수행하는 그리드 컴퓨팅(Grid Computing)에 대한 연구가 최근 활성화 되고 있다. 이에 본 논문에서는 그리드를 구성하는 노드에 작업을 분배하는 메타 스케줄링(Metascheduling)과 노드내의 작업을 프로세서에 적절히 할당하는 작업 스케줄링 방법을, 작업 큐와 두 개의 백필 큐를 사용하는 혼합 기법을 사용하여, 작업 지연율(Slowdown Ratio)을 줄이고 그리드 컴퓨팅의 이용률(Utilization)을 높이는 방법을 제안하였으며 다양한 실험을 통하여 성능을 평가하였다.

  • PDF

Grid Scheduling System for Effective Resource Selection in Service-based Grid Computing (서비스 기반 그리드 컴퓨팅에서의 효과적인 자원 선택을 위한 그리드 스케쥴링 시스템)

  • Hahm Jaegyoon;Kwon Oh-Kyoung;Kim Sangwan
    • Proceedings of the Korean Information Science Society Conference
    • /
    • 2005.07a
    • /
    • pp.592-594
    • /
    • 2005
  • 그리드 스케쥴링 시스템은 그리드 환경에서 매우 다양하고 많은 자원에 대한 서비스들을 통합하여 사용자에게 일관된 접근 방법을 제공하고, 자원들의 동적인 상태나 자원의 서로 다른 정책 등으로 인해 야기되는 복잡성을 사용자에게 감추어 편리한 자원 사용 환경을 제공해 주게 된다 그리드 스케쥴링 문제에서는 기존의 스케쥴링 문제와는 다른 조건들이 존재하는데 자원에 대한 전적인 권한이 스케쥴러에게 주어지지 않는다는 점과 사용자에게 할당 가능한 자원의 양은 풍부하다는 점 등이다. 본 논문에서는 이러한 필요와 조건 하에 설계되고 구현 된 그리드 스케쥴링 서비스에 대해 다룬다. 본 논문에서 설명하는 그리드 스케쥴링 서비스는 그리드 작업 및 자원 관리 시스템에 자원의 발견, 선택, 예약 기능을 제공하고, GGF의 표준인 OGSI 스펙을 따르는 인터페이스를 제공한다. 특히 사용자의 요구조건과 자원의 정책을 만족하는 자원을 적절히 선택해 줌으로써 작업 실행에 있어서의 자동화를 제공하고, 다양한 어플리케이션에 적합한 자원 선택 알고리즘을 추가할 수 있도록 함으로써 특정 어플리케이션을 위한 최적의 자원을 선택할 수 있게 하였다.

  • PDF