• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.3_4
• /
• pp.213-223
• /
• 2004

A dynamic thread pool model is one of a multi-thread server programming model that handles many requests from users concurrently. In most cases, bounded thread pool model is used for server programming. Because it reduces a thread overhead by creating a pool of threads in advance, it ran response more quickly to users' requests. But this model always occupies system resource when there are small amount of requests, which prevents other applications from using available resources. In this paper, for the utilization of occupied but unused resources, we proposed and implemented a prediction-based dynamic thread pool scheme using customized exponential average. From the experiments, we show that this scheme outperforms hounded thread pool model and uses small resources.

• The KIPS Transactions:PartD
• /
• v.11D no.4
• /
• pp.975-982
• /
• 2004

In current P2P concept, pure P2P and Hybrid P2P structures are used commonly. Gnutella and Ktella are forms of pure P2P. and forms of Hybrid P2P are innumerable. File searching models exist in these models. These models provide group management for file sharing, searching and indexing. The general file sharing model is good at maintaining connectivity. However, it is defective in group management. Therefore, this study approaches hierarchical structure in file sharing models through routing technique and backup system. This system was designed so that the user was able to maintain group efficiency and connection reliability in large-scale network.

• The KIPS Transactions:PartA
• /
• v.8A no.4
• /
• pp.299-310
• /
• 2001

In this paper, we study the methodology of distributing the requests of clients based on the state of real server in the LVS(Linux virtual server) clustering system. The WLC(weighted least connection) algorithm in the LVS cluster system is studied. The load distributing algorithm which assigns a weight into each real server is devised by considering the load of real servers. The load test is executed to estimate the load of real servers using a load generating tool. Throughout the result of the experiment, we suggest new load distributing algorithm based on the usage of physical memory of each real server. It is shown that the correction potentiality of new algorithm is somewhat better than that of the WLC algorithm.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.1
• /
• pp.99-114
• /
• 2000

In the secure DNS the amount of information that should be managed greatly increased and the interdependency in the information became very complex. Therefore, it became necessary to develop a mechanism which can manage zone information efficiently. Moreover, a consistent interface became also necessary so that a secure DNS may be efficiently interconnected with other Internet application services. In this paper we explain the design and implementation of a secure DNS extended with two functions : (1) a dynamic update function which enables to add and remove zone information dynamically and (2) a zone transfer function that efficiently transfers update zone information among DNS servers. We developed a method which integrates two zone transfer mechanisms, full zone transfer and incremental zone transfer, and also proposed a method to compress data in the zone transfer message. We also introduced a data structure called a delta file to integrate the zone transfer function and the dynamic update function.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10a
• /
• pp.74-76
• /
• 1998

본 논문에서는 유닉스 파일 시스템에서의 버퍼캐쉬 크기에 따라 사용자 데이터와 메타데이터의 버퍼 캐쉬 히트율을 분석하였다. 그리고 메타 데이터가 유닉스 운영체제 파일 시스템의 성능에 미치는 영향을 분석하고 이를 기반으로 버퍼 캐쉬의 동적 특성과 성능의 장애 요인들을 분석하였다. 유닉스 운영체제에서 사용되는 사용자 데이터와 메타데이터에 대한 버퍼 캐쉬의 동적인 동작을 분석하기 위하여 trace-driven방법을 이용하였으며 이를 위하여 시뮬레이터를 작성.사용하였다. 파일 시스템은 특정 유닉스 버전에 영향을 받지 않기 위해 USF[1]에 기초하였고, 작업부하(workload)로는 Sprite- trace 데이터 중 allspice 서버에서 추출한 데이터를 사용하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.04a
• /
• pp.598-600
• /
• 2004

Mobile Ad hoc Network (MANET)는 자율적으로 작동할 수 있는 멀티 홈 무선 네트워크로서 토폴로지 변화가 자주 일어나며 예측할 수 없는 특성을 갖는다. 특히 scale 이 큰 MANET 환경에서는 더욱더 예측할 수 없는 특성을 갖게 된다. 최근 MANET에서의 라우팅을 하기 위한 주소 할당에 관심이 모아지고 있는 상황에서 각 노드는 DHCP 같은 서버의 역할 없이 주소를 할당할 수 있는 기능을 가져야 하며, 이동 단말들이 모두 IP stack을 지원하는 것을 감안해서 IP주소를 사용해야할 것이다. 따라서 본 논문에서는 scale이 큰 MANET 환경에서 동적으로 IP 주소를 할당하는 방법을 제안한다.

• Proceedings of the Korean Information Science Society Conference
• /
• 1999.10a
• /
• pp.15-17
• /
• 1999

본 논문은 다수의 사용자에 대한 실시간 멀티미디어 서비스에서 문제가 되는 초기 지연시간을 최소화하는 새로운 동적 스트림 합병 기법을 제안한다. 제안된 기법은 멀티미디어 서비스, 특히 비디오 서비스의 경우 약간의 QoS 변화가 서비스의 질에 큰 영향을 미치는 않는 점을 이용하여, 시간적으로 서로 인접한 여러 개의 스트림을 점차적으로 합병하여 서비스함으로써 짧은 초기 지연시간 유지를 위한 버퍼의 양을 최소화 할 수 있다. 성능분석 결과에서 제안된 기법은 기존의 방법들에 비하여 버퍼 활용면에서 우수한 성능을 나타냄을 보인다.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.12
• /
• pp.3652-3661
• /
• 1999

To support multimedia applications such as a multimedia communication systems and multimedia broadcasting, an operating system need to predict their timing-constraints. So, In this real-time systems, we must solve the priority inversion problem that may make the behavior of unpredictable systems and need a real-time server model that provides a better preemptability and minimizes a system overhead. In current real-time systems, the single thread server model, the worker model and the dynamic server model are being used for synchronization but they cannot propose an effective structure for managing shared resources. In this paper, the priority inheritance protocol is used to solve the priority inversion problem and the hybrid prioritized worker model is proposed, which can provide a more effective structure and a faster response time minimizing a system overhead. The hybrid prioritized worker model is to combine the static and the dynamic prioritized worker model, and have a better performance than other models in response time which is an important factor in a real-time system.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.4
• /
• pp.79-85
• /
• 2013

In SIP(Session Initiation Protocol) environment, the presence server should process SIP SUBSCRIBE request messages including multiple presence resources addresses from users, and also precess massive notification messages from the subscribed presence resources. The load of the presence server increases as number of users increase, and it limits system extendability. Therefore a new multiple presence servers architecture has been suggested in this research. In this architecture presence servers can be added dynamically and each server's load can be controlled effectively as number of users increase. Each presence server can monitor current load status of entire presence system by using presence event notification package which newly has been suggested in this paper. When a particular presence server's load increases over predefined limit, the presence service processing is distributed by selecting a server which has the smallest load, or by generating a new server dynamically. In this system the overall load of the entire system can be controlled optimally and extendability of the system can be increased. For this purpose a new presence event notification package and presence information data format have been suggested. The performance of the proposed system has been evaluated by experiments. They shows 44.3% increase in SUBSCRIBE message processing time, and 43.1% increase in Notification message processing time.

• Proceedings of the Korean Information Science Society Conference
• /
• 2000.04b
• /
• pp.726-728
• /
• 2000

본 논문에서는 최근 인터넷 상에서 표준 공통 포맷으로 대두되고 있는 XML을 이용하여 웹 기반 원격 교육 시스템에서 강의 내용에 참조도리 참고문헌 지원 시스템을 설계하고 구현하였다. Three-tier 환경에서 구현한 이 시스템은 middle-tier인 웹 서버에서 데이터베이스에 저장된 참고문헌을 XML 데이터로 변환하여 효과적으로 처리함으로써 서버의 부하를 감소시키며 이것은 성능 향상으로 이어져 학습자에게 더 나은 속도로 원격 교육의 참고문헌 정보서비스를 제공할 수 있다. 또한 동적으로 서버와 상호작용 가능하도록 학습자가 c마고 문헌의 유익함 정도를 매긴 등급 점수 계산에 직접 참여시켜 그 결과를 볼 수 있도록 하여 학습 의욕을 더욱 고취시킬 수 있다. 앞으로 웹기반 원격교육의 참고문헌 지원 시스템은 세계 각 대학이나 연구소에 분산되어 있는 여러 데이터 소스로부터 필요한 정보만을 실시간으로 추출하여 수집, 통합, 통계 처리할 수 있도록 확장될 수 있을 것이다.