• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2017.05a
• /
• pp.103-105
• /
• 2017

Hadoop MapReduce is a processing framework in which users' job can be efficiently processed in parallel and distributed ways on the Hadoop cluster. MapReduce task schedulers are used to select target nodes and assigns user's tasks to them. Previous schedulers cannot fully utilize resources of Hadoop cluster because they does not consider dynamic characteristics of cluster based on nodes' availability. To increase utilization of Hadoop cluster, this paper proposes a novel task assignment policy for MapReduce that assigns a job tasks to dynamic cluster efficiently by considering availability of each node.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10b
• /
• pp.111-113
• /
• 1998

공간조인은 지리정보 시스템에서 공간분석을 위한 주요 연산중의 하나이다. 이러한 공간조인은 대상이 되는 공간 객체의 수가 증가함에 따라 연산시간이 지수적으로 증가하는 특징을 가지고 있다. 그래서 대규모 공간 데이터에 다한 공간 연산시간을 줄이기 위한 처리기법이 연구되고 있다. 그렇지만, 공유 디스크 구조에서 다중 프로세서의 디스크 동시 접근으로 인한 병목현상을 완화하고, 프로세서간의 공유 디스크 구조에서 다중 프로세서의 디스크 동시 접근으로 인한 병목현상을 완화하고, 프로세서간의 메시지 전달을 최소화하기 위한 태스크 생성방법, 태스크 할당방법에 관한 구체적인 연구가 없었다. 그래서 우선 병렬 공간 조인의 성능저하 요인을 분석하고, 이에 대한 성능 향상방안을 제시한다. 구체적으로 디스크 접근 시간을 줄이기 위한 객체 캐쉬 방법과 시공간 지역성을 이용한 태스크 생성 및 할당방법을 제시한다. 그리고 제안한 방법들에 대해 실험평가를 통해 최대 7.2배의 성능증가를 획득할 수 있음을 보여준다.

• Journal of KIISE:Computer Systems and Theory
• /
• v.34 no.8
• /
• pp.337-347
• /
• 2007

An embedded system is called a multi-mode embedded system if it performs multiple applications by dynamically reconfiguring the system functionality. Further, the embedded system is called a multi-mode multi-task embedded system if it additionally supports multiple tasks to be executed in a mode. In this Paper, we address a HW/SW partitioning problem, that is, HW/SW partitioning of multi-mode multi-task embedded applications with timing constraints of tasks. The objective of the optimization problem is to find a minimal total system cost of allocation/mapping of processing resources to functional modules in tasks together with a schedule that satisfies the timing constraints. The key success of solving the problem is closely related to the degree of the amount of utilization of the potential parallelism among the executions of modules. However, due to an inherently excessively large search space of the parallelism, and to make the task of schedulabilty analysis easy, the prior HW/SW partitioning methods have not been able to fully exploit the potential parallel execution of modules. To overcome the limitation, we propose a set of comprehensive HW/SW partitioning techniques which solve the three subproblems of the partitioning problem simultaneously: (1) allocation of processing resources, (2) mapping the processing resources to the modules in tasks, and (3) determining an execution schedule of modules. Specifically, based on a precise measurement on the parallel execution and schedulability of modules, we develop a stepwise refinement partitioning technique for single-mode multi-task applications. The proposed techniques is then extended to solve the HW/SW partitioning problem of multi-mode multi-task applications. From experiments with a set of real-life applications, it is shown that the proposed techniques are able to reduce the implementation cost by 19.0% and 17.0% for single- and multi-mode multi-task applications over that by the conventional method, respectively.

• Journal of KIISE:Computer Systems and Theory
• /
• v.33 no.1_2
• /
• pp.1-14
• /
• 2006

Sort-last parallel rendering using a cluster of GPUs has been widely used as an efficient method for visualizing large- scale volume datasets. The performance of this method is constrained by load balancing when data parallelism is included. In previous works static partitioning could lead to self-balance when only task level parallelism is included. In this paper, we present a load balancing scheme that adapts to the characteristic of volume dataset when data parallelism is also employed. We effectively combine the hierarchical data structures (octree and BSP tree) in order to skip empty regions and distribute workload to corresponding rendering nodes. Moreover, we also exploit a 3D clustering method to determine visibility order and save the AGP bandwidths on each rendering node. Experimental results show that our scheme can achieve significant performance gains compared with traditional static load distribution schemes.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10c
• /
• pp.633-635
• /
• 2000

본 논문에서는 프로그램을 루프 구조에 근거하여 계층적으로 표현한 HTG (Hierarchical Task Graph)의 복합 노드 태스크들을 공유 메모리 다중처리기 환경에서의 효율적 수행을 위한 새로운 스케쥴링 기법을 제안한다. 단일처리기의 멀티스레드 구조를 비롯한 여러 플랫폼에 적용하기 위해 자바의 스레드를 사용하여 구현하였으며, 기존의 HTG의 함수 병렬성을 위한 비티 벡터 알고리즘과 성능을 비교 분석하였다. 실험 결과에서 보듯이, 제안된 기법이 비트 벡트 방법에 비해 수행 시간 측면에서 효율적임을 알 수 있으며 또한, 좋은 부하 균형을 유지하였다.

• Journal of KIISE:Computer Systems and Theory
• /
• v.35 no.9_10
• /
• pp.429-440
• /
• 2008

As embedded system becomes more complex, software development becomes more important in the entire design process. Most embedded applications consist of multi -tasks, that are executed in parallel. So, dataflow model that expresses concurrency naturally is preferred than sequential programming language to develop multitask software. For the execution of multitasking codes, operating system is essential to schedule multi-tasks and to deal with the communication between tasks. But, it is needed to execute multitasking code without as when the target hardware platform cannot execute as or target platforms are candidates of design space exploration, because it is very costly to port as for all candidate platforms of DSE. For this reason, we propose the serialized multitasking code generation technique from dataflow specification. In the proposed technique, a task is specified with dataflow model, and generated as a C code. Code generation consists of two steps: First, a block in a task is generated as a separate function. Second, generated functions are scheduled by a multitasking scheduler that is also generated automatically. To make it easy to write customized scheduler manually, the data structure and information of each task are defined. With the preliminary experiment of DivX player, it is confirmed that the generated code from the proposed framework is efficiently and correctly executed on the target system.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.10c
• /
• pp.621-623
• /
• 2000

클러스터를 구성하는 단위 노드로 SMP가 새로운 대안으로 시장에 등장하였다. 이러한 멀티프로세서 클러스터(CLUMP)는 하나의 시스템에 다중 메모리 구조를 가지는데, CLUMP가 가지는 다중 메모리 구조를 효과적으로 사용하기 위해서 본 논문에서는 중첩된 병렬화 프로그램 모델을 제안하였다. 중첩된 병렬화 모델은 중첩된 루프 레벨의 병렬화, 중첩된 태스크 레벨의 병렬화, 그리고 다중 중첩된 병렬화로 나뉜다. 본 논문에서는 중첩된 루프 레벨의 병렬화를 실험대상으로 하여 그 성능을 평가하고 단일 메모리 구조의 병렬화 프로그램과 성능을 비교하였다. 실험 결과 시험한 중첩된 병렬화 모델이 단일 메모리 구조의 병렬화 프로그램에 비하여 좋은 성능을 나타내었지만, 실험대상이 된 루프 레벨 병렬화의 잠재적인 특징으로 인해 실행에 참여하는 노드 수가 많아질수록 성능 향상 폭이 감소하는 결과를 보였다. 프로그램의 성능 향상 폭과 확장성은 문제 크기가 클수록 좋은 특성을 보였다.

• The KIPS Transactions:PartD
• /
• v.8D no.4
• /
• pp.347-360
• /
• 2001

The most expensive spatial operation in spatial databases is a spatial join which computes a combined table of which tuple consists of two tuples of the two tables satisfying a spatial predicate. Although the execution time of sequential processing of a spatial join has been so far considerably improved, the response time is not tolerable because of not meeting the requirements of interactive users. It is usually appropriate to use parallel processing to improve the performance of spatial join processing. However, as the number of processors increases, the efficiency of each processor decreases rapidly because of the disk bottleneck and the overhead of message passing. This paper proposes the method of task allocation to soften the disk bottleneck caused by accessing the shared disk at the same time, and to minimize message passing among processors. In order to evaluate the performance of the proposed method in terms of the number of disk accesses and message passing, we conduct experiments on the two kinds of parallel spatial join algorithms. The experimental tests on the MIMD parallel machine with shared disks show that the proposed semi-dynamic task allocation method outperforms the static and dynamic task allocation methods.

• Journal of KIISE:Computer Systems and Theory
• /
• v.27 no.3
• /
• pp.264-276
• /
• 2000

Recent advances of technologies make easy sharing various information and utilizing system resources on the Internet. Especially, code migration using applets of Java supports the distribution of programs on the web environment, and also browsers executing the applets guarantee the reliability of a migrated codes. In this paper, we describe the design and implementation of a web-based parallel processing system, which distributes migratable codes of a large job, makes the distributed codes to execute in parallel, and controls and gathers the results of each execution. The hosts participate in the computation reside on the Internet, spreaded out geographically, and the heterogeneity and the variability among them are severe. Thus, task allocation considering the performance differences and the adaptability to the severe variability are necessary. We present an adaptive task allocation algorithm applied to our system and the performance evaluation.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2016.05a
• /
• pp.811-812
• /
• 2016

최근 공간 정보는 생산 되는 양과 데이터의 생성 빈도 및 다양성으로 인해 기존의 공간 데이터베이스 시스템에서 처리하기 어렵다. 그래서 공간 정보는 빅데이터와 연계에 관한 시도가 활발히 진행되고 있다. 그러나 효율적인 단일할당, 다중할당 색인기반 공간 연산에 대한 연구는 거의 없다. 이 논문에서는 공간 연산 중 비용이 매우 큰 공간 조인을 빅데이터 시스템에서 처리하기 위한 고려요소를 제시하고자 한다. 구체적으로 맵리듀스 시스템의 태스크 할당을 위한 단일 할당 공간 색인방안을 설명하고, 불균일 분포가 심한 공간 정보의 특성을 고려한 부하 균등화 시 고려 요소를 제시하고자 한다. 맵리듀스와 같은 병렬 공간 데이터베이스 시스템에서의 두 가지 문제인 데이터 불균일 분포 문제와 경계 겹침 색인의 문제와의 연관성을 기술한다.