• Title/Summary/Keyword: processing load

Search Result 1,477, Processing Time 0.032 seconds

Performance Counter Analysis for Effective Adaptive Load Balancing (효과적인 Adaptive Load Balancing을 위한 성능 지표 분석)

  • Lim, Yoo-Jin;Lee, Won-Q;Han, Young-Tae;Lee, Dong-Hoon;Choi, Eun-Mi
    • Proceedings of the Korea Information Processing Society Conference
    • /
    • 2002.11a
    • /
    • pp.143-146
    • /
    • 2002
  • 웹 서비스를 제공하는 분산 서버 시스템에서 각 서버의 부하 상태를 파악하여 처리해야 할 부하를 조절하여 주면 서버의 부하가 균등하게 되어 더 나은 성능을 얻을 수 있게 된다. 서버의 부하 상태는 시스템의 자원에 영향을 미치는 여러 가지 요소에 의하여 분석을 할 수 있다. 본 논문에서는 다양한 스트레스 테스트를 통하여 서버의 자원의 고갈을 나타내는 주요 성능 지표들을 변화 상태를 분석하였다. 고려된 성능지표로는 Available Memory 양, Page Read 수, Processor Utilization, Processor Queue Length, 네트웍으로 전달된 Transmitted Bytes, 연결된 Connection 개수이다. 실제로 이중 하나의 요소를 적용시켜서 ALBM (Adaptive Load Balancing Mechanism)을 실행을 하였을 때 일반 LVS Round Robin 보다 성능이 좋은 결과를 낳았다.

  • PDF

A countermeasure against Foreshadow and ZombieLoad attacks based on segmentation fault monitoring (Segmentation fault 모니터링을 통한 Foreshadow 및 ZombieLoad 공격 방어 기법 연구)

  • Lee, Jun-Yeon;Suh, Taeweon
    • Proceedings of the Korea Information Processing Society Conference
    • /
    • 2019.10a
    • /
    • pp.384-387
    • /
    • 2019
  • 2018년 Meltdown 공격이 발표된 이후 Foreshadow, ZombieLoad 등 다양한 종류의 마이크로아키텍처 기반 부 채널 공격과 방어 기법들이 발표되었다. 그중 Meltdown 공격을 원천 차단할 수 있는 KPTI (Kernel Page Table Isolation)는 커널 영역을 사용자 메모리 영역과 분리하여 커널 정보의 유출을 방어할 수 있으나, 최대 46%의 시스템 성능 저하를 가져온다. 본 연구는 런타임에 시스템콜 발생빈도를 분석하여 낮은 오버헤드로 Meltdown-type 공격을 탐지하고, 방어하는 프로그램을 개발하고 실험하였다. 개발한 Fault Monitoring Tool은 기존 시스템 대비 적은 오버헤드(최대 7%)로 악의적인 사용자를 구분해 내고 방어할 수 있었다.

A CRL Distribution Scheme Minimizing the Time for CRL Processing of Vehicles on Vehicular Communications

  • Kim, Hyun-Gon
    • Journal of the Korea Society of Computer and Information
    • /
    • v.23 no.12
    • /
    • pp.73-80
    • /
    • 2018
  • Certification revocation list(CRL) is needed for excluding compromised, faulty, illegitimate vehicle nodes and preventing the use of compromised cryptographic materials in vehicular communications. It should be distributed to vehicles resource-efficiently and CRL computational load of vehicles should not impact on life-critical applications with delay sensitive nature such as the pre-crash sensing that affords under 50msec latency. However, in the existing scheme, when a vehicle receives CRL, the vehicle calculates linkage values from linkage seeds, which results in heavy computational load. This paper proposes, a new CRL distribution scheme is proposed, which minimizes the time for CRL processing of vehicles. In the proposed scheme, the linkage value calculation procedure is performed by road-side unit(RSU) instead of the vehicle, and then the extracted linkage values are relayed to the vehicle transparently. The simulation results show that the proposed scheme reduces the CRL computational load dramatically, which would minimize impact on life-critical applications' operations with low latency.

Multipath Routing for Load Balancing in Wireless Mesh Network

  • Gao, Hui;Kwag, Young-wan;Lee, Hyung-ok;Nam, Ji-seung
    • Proceedings of the Korea Information Processing Society Conference
    • /
    • 2013.11a
    • /
    • pp.548-550
    • /
    • 2013
  • Multipath routing is proposed to perform better load balancing and to provide high fault tolerance. In this paper common factors on multi-path routing for WMNs are discussed and joint multi-channel and multipath routing approaches for load balancing are introduced at the same time. The key issues on how to design an effective multi-path routing protocol with appropriate performance metrics for WMNs are also discussed through the case studies in the paper.

A Study on Load Balancing Method using Dual Structure Queue Circulation on Clustering (클러스터링에서의 이중 구조 큐 순환을 이용한 부하 분산 기법에 관한 연구)

  • Ku, Min;Min, Dugki
    • Proceedings of the Korea Information Processing Society Conference
    • /
    • 2009.11a
    • /
    • pp.117-118
    • /
    • 2009
  • 클러스터링 분야에서는 대규모 계산 및 처리, 저장 등의 작업을 하기 위해 다양한 정책과 효율적인 전략들이 기술적 사상으로써 적용되는 대표적인 분야 중 하나이다. 그 중에서도 대규모 처리(Load)시 처리자원으로의 효율적인 자원 분배(Balancing)를 하는 LoadBalancing에 관한 연구는 지속적으로 이루어지고 있다. 효율적인 로드밸런싱 기법에는 경제적인 계산 비용이 소요되는 서버풀내의 각 처리 서버들에 대한 연결 정보를 Queue에 넣고 FCFS 스케줄링 알고리즘등을 활용하는 방법들이 널리 쓰이고 있다. 그러나, 이 방법의 경우, 서버풀의 추가 및 삭제와 같은 변동상황 발생시 이를 처리하는데 많은 비용이 소요된다는 문제를 가지고 있다. 본 논문에서는 클러스터링 처리 환경에서 이중 구조 큐 순환을 이용한 서버풀 및 이에 속하는 각각의 처리서버에 대한에 부하 분산 기법을 소개하고자 한다.

Signal Processing of Capacitive Load and Gap Measurement with High Precision Using Surface Acoustic Wave Device (표면 탄성파 장치를 이용한 용량성 부하의 신호처리 및 이를 이용한 초정밀 간극 측정)

  • Kim, Jae-Geun;Lee, Taek-Joo;Lim, Soo-Cheol;Park, No-Cheol;Park, Young-Pil;Park, Kyoung-Soo
    • Proceedings of the Korean Society for Noise and Vibration Engineering Conference
    • /
    • 2009.10a
    • /
    • pp.376-380
    • /
    • 2009
  • Surface acoustic wave (SAW) device is widely used as a bandpass filter, a chemical or physical sensor, and an actuator. In this paper, we propose the capacitive gap measurement system with high precision through the signal processing using SAW device. The research process is mainly composed of theoretical part and experimental part. In theoretical part, equivalent circuit model was used to simulate the SAW response by the change of capacitance. In experimental part, commercialized capacitor was used to see the SAW response by the change of load capacitance. After that, gap adjustment system was made physically and the SAW response by the change of gap which caused the capacitance change was measured. And resolution and stroke was decided comparing the signal change and basic measurement noise level.

  • PDF

Effective Destination Decision Policies for Dynamic Load Balancing in Distributed Computer Systems with Star Topology (스타형 분산 컴퓨터 시스템의 동작 부하분산을 위한 효율적인 위치결정 정책)

  • Im, Gyeong-Su;Ha, Seong-Ryong;Kim, Jong-Geun
    • The Transactions of the Korea Information Processing Society
    • /
    • v.2 no.5
    • /
    • pp.768-776
    • /
    • 1995
  • In a distributed computer system with star topology, the processing power of central node may be greater than that of peripheral nodes and the job arrival rate of each node may be different. The performance of load balancing may be very different according to the selection of sender criteria and receiver criteria for job transfer. But the optimal decision of sender and receiver criteria is very difficult in heterogeneous distributed computer systems. In this paper we propose effective destination decision policies using dynamic criteria such as mean job response time of a node and estimated response time of a job, instead of fixed threshold for dynamic load balancing.

  • PDF

Low-grade waste heat recovery and repurposing to reduce the load on cooling towers

  • McLean, Shannon H.;Chenier, Jeff;Muinonen, Sari;Laamanen, Corey A.;Scott, John A.
    • Advances in Energy Research
    • /
    • v.7 no.2
    • /
    • pp.147-166
    • /
    • 2020
  • Industrial cooling towers are often ageing infrastructure that is expensive to maintain and operate. A novel approach is introduced in which a heat pump circuit is incorporated to reduce the load upon the towers by extracting low-grade energy from the stream sent to the towers and repurposing in on-site processing operations. To demonstrate the concept, a model was constructed, which uses industrial data on cooling towers linked to a smelter's sulphuric acid plant, to allow direct economic and environmental impact comparison between different heat recovery and repurposing scenarios. The model's results showed that implementing a heat pump system would significantly decrease annual operating costs and achieve a payback period of 3 years. In addition, overall CO2 emissions could be reduced by 42% (430,000 kg/year) and a 5% heat load reduction on the cooling towers achieved. The concept is significant as the outcomes introduce a new way for energy intensive industrial sectors, such as mineral processing, to reduce energy consumption and improve long-term sustainable performance.

Indivisible load scheduling applied to Linear Programming (선형계획법을 적용한 임의 분할 불가능한 부하 분배계획)

  • Son, Kyung-Ho;Lee, Dal-Ho;Kim, Hyoung-Joog
    • 한국정보통신설비학회:학술대회논문집
    • /
    • 2005.08a
    • /
    • pp.382-387
    • /
    • 2005
  • There are many studies on arbitrarily divisible load scheduling problem in a distributed computing network consisting of processors interconnected through communication links. It is not efficient to arbitrarily distribute the load that comes into the system. In this paper, how to schedule in case that arbitrarily indivisible load comes into the system is studied. Also, the cases of the divisible load mixed with the indivisible load that come into network were dealt with optimal load distribution in parallel processing system by scheduling applied to linear programming.

  • PDF

Silicon Strain Gauge Load Cell for Weighting Disdrometer

  • Lee, Seon-Gil;Moon, Young-Soon;Son, Won-Ho;Sohn, Young-Ho;Choi, Sie-Young
    • Journal of Sensor Science and Technology
    • /
    • v.22 no.5
    • /
    • pp.321-326
    • /
    • 2013
  • In this paper, the usability of a compact silicon strain gauge load cell in a weighting disdrometer for measuring the impact load of a falling raindrop is introduced for application in a multi-meteorological sensor. The silicon strain gauge load cell is based on the piezoresistive effect, which has a high linearity output from the momentum of the raindrop and the simplicity of signal processing. The weighting disdrometer shows a high sensitivity of 7.8 mV/g in static load measurement when the diaphragm thickness of the load cell is $250{\mu}m$.