QoS 기반 가상 머신 태스크 스케줄링

장준혁(Joonhyouk Jang),정진만(Jinman Jung),홍지만(Jiman Hong) 한국정보과학회 2021 정보과학회 컴퓨팅의 실제 논문지 Vol.27 No.7

가상화 플랫폼은 인터럽트 발생을 인지하고 가상 머신으로 인터럽트를 전달함으로써 I/O를 가상화하고 QoS(Quality of Service)를 높이기 위해 I/O 인터럽트가 전달된 가상 머신을 우선 처리하는 스케줄링 정책을 사용한다. 그러나 가상 머신 스케줄러가 I/O 인터럽트 처리를 우선하는 경우, 특정 가상 머신이 처리기 자원을 독점하거나 실시간 응용을 실행하는 가상 머신들이 경쟁 상태에 놓이는 상황이 발생할 수 있다. 따라서 스케줄링의 공정성과 응답 시간의 트레이드오프(trade-off)를 고려한 가상 머신 스케줄링 기법이 요구된다. 본 논문에서는 가상 머신의 자원 소모율을 통해 실행 패턴을 예측하고, 예측값에 따라 스케줄링 우선순위를 동적으로 제어하는 QoS 기반 가상 머신 스케줄링 기법을 제안한다. 성능 평가를 통해 제안 기법이 적은 스케줄링 오버헤드로 응답성과 공정성을 보장함을 보였다. In a virtualization platform, I/O interrupt is propagated to a corresponding virtual machine through a hypervisor. A virtual machine scheduler in the hypervisor preferentially schedules the virtual machine. This mechanism enhances the quality of service in the I/O virtualization in common cases. However, the mechanism has hazards such that a skewed workload can cause monopolization or a race condition in CPU resources. Therefore, a hypervisor requires a virtual machine scheduler that considers the trade-off between fairness and response time in virtual machine scheduling. In this paper, we propose a fair and responsive virtual machine scheduler. The proposed scheme dynamically adapts the priorities of virtual machines based on provisions on the execution patterns of the virtual machines. We evaluated the proposed scheme with simulations. The result showed that the proposed scheme provides a fair and responsive scheduling algorithm with low computational overheads.

Process Monitoring and Part Program Optimization Using Virtual Machine Tools

김창주(Chang-Ju Kim),허세곤(Segon Heo),이찬영(Chan-Young Lee),오정석(Jung Seok Oh) Korean Society for Precision Engineering 2022 한국정밀공학회지 Vol.39 No.12

A virtual machine tool, a computer simulation model of the machine motion and cutting process with a level of accuracy and consistency that can replace an accurate machine tool, is one of the critical digital transformation technologies in the manufacturing industry. During the machine development phase, cost and time can be reduced by evaluating machining efficiency and quality through virtual prototyping. In the machine application phase, virtual machine tools can be used to accurately assess the condition of equipment and processes by analyzing actual data combined with simulated data. This paper introduces a virtual machine tool system that can analyze the behavior of an accurate machine tool by integrating physical models of structure, numerical controller, and cutting process. The key features of the virtual machine tool, synchronous machining simulation, machining stability detection, machining error estimation, and part program optimization, were evaluated through various machining tests with a vertical 3-axis milling machine.

SaaS 기반 이동형 개인 맞춤 소프트웨어 플랫폼을 위한 VM의 Host OS와 Guest OS의 네트워크 성능 측정 방법 개선

우수정 ( Sujeong U ),온진호 ( Jin-ho On ),최정란 ( Jung-rhan Choi ),최완 ( Wan Choi ),이문근 ( Moon-kun Lee ) 한국인터넷정보학회 2009 인터넷정보학회논문지 Vol.10 No.2

최근 들어 SaaS기반 이동형 개인 맞춤 소프트웨어 플랫폼에 관한 연구 및 개발이 활발해지고 있다. 이러한 플랫폼들은 다양한 사용자의 OS를 만족시키기 위해 최적화된 가상머신이 필수적으로 요구된다. 또한 다양한 Host OS에서 내부, 외부 네트워크간에 빠르고 안정적인 서비스를 지원 함으로서 사용자의 작업환경 이동성을 보장해야 한다. 이를 위해 가상머신은 다양한 관점에서의 성능 측정이 필요하다. 하지만 기존 연구에서 가상머신의 성능 측정은 Host 컴퓨터에 설치된 VM에 Guest OS를 설치하여 이를 하나의 컴퓨터로 간주하고, 외부의 클라이언트에서 네트워크 성능을 측정한다. 이는 이동형 개인 맞춤 소프트웨어플랫폼을 위한 가상머신 성능 측정에 적합하지 않다. 본 논문은 SaaS기반 이동형 개인 맞춤 소프트웨어 플랫폼을 위한 최적화된 가상머신을 선정하기 위해서 네트워크 성능 측정 방법과 측정된 결과 분석을 통한 최적화된 가상머신을 제안한다. Recently, there are a number of researches and developments for the personalized software platform for mobility based on SaaS. The platform requires an optimal virtual machine in order to satisfy the operating systems of various users for the software. In addition, the platform must guarantee the mobility of the users` working environments by supporting fast and secure services between internal and external networks in the platform operating systems. In order to verify the optimal behaviors of virtual machines for the platform, the performance of the virtual machines must be measured and analyzed in various perspectives. In the previous research, unfortunately, the performance of a virtual machine were conducted in the condition that a guest operating system was installed on the virtual machine and considered as a computer, by measuring the network traffic between the guest operating system and an external client operating system. This performance measurement was not suitable for a virtual machine for the platform since a number of different software must be handled in the virtual machine. In order to overcome this limitation, this paper presents a measurement method for network performance and proposes the most optimal virtual machine by the method.

가상화 환경에서 동적 임계치 기반 가상 머신 이주 기법

최호근 ( Ho Gun Choi ),박지수 ( Ji Su Park ),손진곤 ( Jin Gon Shon ) 한국컴퓨터교육학회 2015 컴퓨터교육학회 논문지 Vol.18 No.2

가상화 환경에서는 물리적 자원을 여러 가상 머신이 같이 사용한다. 그러나 특정 가상 머신이 컴퓨팅 자원을 많이 쓰면 다른 가상 머신들이 동작하지 못하게 된다. 이러한 문제를 해결하기 위한 다양한 방법이 있다. 이 중 대표적인 것이 특정의 가상 머신을 다른 서버(이 서버를 타겟 서버라고 함)에 이주시키는 방법이다. 이는 가상 머신을 타겟 서버에 이주시키면서 서버의 과부하가 전이되는 현상이 있고, 가상머신을 다시 다른 서버로 이주시켜야 하는 문제점이 있다. 본 논문에서는 이러한 문제점을 해결하기 위해 동적으로 임계치를 적용하여 이주 대상을 결정하는 알고리즘을 제안한다. 동적 임계치를 적용한 이주 알고리즘은 다음과 같은 특징을 가진다. 첫째, 서버의 CPU, 네트워크, 그리고 메모리 등의 자원 사용률의 변화에 따라 동적으로 임계치를 적용한다. 둘째, 서버에서 임계치를 초과한 자원을 기준으로 가상 머신 집합과 타겟 서버를 결정한다. 셋째, 타겟 서버의 자원 사용률을 기준으로 가상 머신을 결정한다. In an virtualization environment, several virtual machines use physical resources together. If a specific virtual machine uses to much of the computing resources, other machines may not be working properly. Thereare various method to solve this problem. Most representative study is to migrate a specified virtual machines to a different server, a target server. In this study, server load can be transferred to a target server by the remigrate of the load imposed on virtual machine. It is still problematic that virtual machine has to remigrate to a different server. This thesis has proposed the algorithm determining the remigration targets by applying dynamic thresholds to solve those problems. The migration algorithm applies dynamic thresholds according to the following criteria. Firstly, the usage of CPU, network and memory; secondly, decide the set of artificial machine and the target server based on the resources surpassed thresholds; thirdly, determine artificial machines based on the resource usage in the target server.

A New Remote USB Architecture for Live Migration of Virtual Machines in SoD (System on-Demand) Service

Dong-oh Kang,Jinho Yoo,Jeunwoo Lee 제어로봇시스템학회 2010 제어로봇시스템학회 국제학술대회 논문집 Vol.2010 No.10

In this paper, we propose a new remote USB architecture to support live migration of virtual machines in SoD (System on-Demand) service. The SoD (System on-Demand) service is the service that provides virtual personal computers which are optimized virtual machines using distributed u-computing devices around users when the users demand. The virtual USB technology for virtual machines is a kind of I/O virtualization to make users of virtual machines use remote USB devices as local USB devices. In a SoD system, a virtual machine can migrate among some SoD servers even when users use virtual machines due to resource reallocation mechanism of the system. In case of the live migration of the virtual machine, the connection of remote USB to the virtual machine will be broken, which means the break of I/O virtualization. For the seamless I/O virtualization of remote USB in spite of live migration of virtual machines, we propose a new remote USB architecture by adopt the virtual connection manager which sustains the virtual connection among the virtual machine and remote USB devices. With the proposed remote USB architecture, users of the virtual machine use remote USB devices seamlessly without noticing live migration of virtual machines.

Design of an Efficient Method for Identifying Virtual Machines Compatible with Service Chain in a Virtual Network Environment

Hyeonseok Oh,Daeun Yu,Yoon-Ho Choi,Namgi Kim 보안공학연구지원센터 2014 International Journal of Multimedia and Ubiquitous Vol.9 No.11

With advancements in network technologies, network virtualization has been proposed to efficiently provide a large number of services and flexible management by utilizing limited resources over existing networks as much as possible. Network virtualization has been proposed as a new paradigm for networks, as it simplifies complicated network configurations for convenient maintenance through the maximum utilization of limited network resources. However, problems have arisen due to network virtualization, such as how to assign real network components that are compatible with virtual network components over a large number of virtual machines existing in a network. In particular, the service-chaining concept—where a network flow only passes through needed services—has been newly introduced by combining the recently highlighted Software-Defined Network with a virtualization concept called Network Function Virtualization. As a result, studies on which virtual machines are selected and how to connect them have increased ever more. Accordingly, this paper aims to identify virtual machines that are compatible with service chaining in a virtual network environment where virtual machines are dispersed, and it proposes a method of how to create a path by connecting virtual machines.

가상머신 상의 PSTR모델 수행환경연구

유진호(Jinho Yoo),한규종(Kyujong Han),김용현(Yong-Hyun Kim),안미림(Mirim Ahn),김두현(Doo-Hyun Kim) 한국정보기술학회 2013 한국정보기술학회논문지 Vol.11 No.8

This paper is an execution environment study to implement the application process based on a PSTR model on a virtual machine. We will load one hardware system with two virtual machines for implementing the PSTR model. We conduct a research about configuration of management module, sharing of serial device simultaneously and communication between tasks on different virtual machines on virtual machine environment for implementing the PSTR model. The management module can be located on the host operating system or on another virtual machine. We decide the position of management module after the comparison of their performances. We analyze the task execution environment related to serial ports on the virtual machine because application tasks hold their serial ports in this study. After analyzing, we design and implement the modified virtual serial port suitable for the PSTR model. And we also make inter task communication functions on different virtual machines and compare the performance of this communication functions with a legacy communication method with their experimental results.

Simulation-based machining condition optimization for machine tool energy consumption reduction

Lee, Wonkyun,Kim, Seong Hyeon,Park, Jaesang,Min, Byung-Kwon Elsevier 2017 JOURNAL OF CLEANER PRODUCTION Vol.150 No.-

<P><B>Abstract</B></P> <P>Optimizing the machining condition is one of the effective ways for reducing the energy consumption of machine tools at a unit process level. Based on statistical approaches with design of experiments, various methods have been developed to reduce the energy consumption by optimizing the machining condition. However, the methods cannot be easily utilized when the optimization target or machine tool design is modified because the optimal solution is determined based on the experimentally measured data. In this study, a simulation-based method that utilizes a virtual machine tool (VMT) to optimize the machining condition is proposed. The VMT model is designed to focus on estimating the energy consumption during machining and is developed by replicating real machine tools. Based on the VMT model, a genetic algorithm is used to optimize the machining condition to reduce the energy consumption. The changes in the optimization target or machine tool design are easily considered by modifying the cost function or component model, respectively. The proposed method is applied to reduce the energy consumption of a three-axis milling machine. The optimal feed rate and spindle speed are obtained for each line of the part program when the thrust force is limited. An experimental setup of the machine tool with an energy consumption monitoring system is constructed to demonstrate the effectiveness of the proposed method. The results show that the total energy consumption of the machine tool reduces by 13% owing to the optimization.</P> <P><B>Highlights</B></P> <P> <UL> <LI> Simulation-based method to reduce the energy consumption of machine tool is proposed. </LI> <LI> Virtual machine tool to estimate the energy consumption during machining is built. </LI> <LI> Optimal machining conditions to minimize the energy consumption is obtained. </LI> <LI> Energy consumption reduction using proposed method is demonstrated experimentally. </LI> </UL> </P>

달빅 가상 머신에서의 장기 추적 방식 적시 컴파일

오형석(Hyeong-Seok Oh),문수묵(Soo-Mook Moon) 한국정보과학회 2011 정보과학회논문지 : 시스템 및 이론 Vol.38 No.3

구글의 모바일 플랫폼인 안드로이드에서는 기존의 자바 어플리케이션을 수행하기 위해 자바 클래스 파일을 덱스(dex) 파일로 변경한 뒤 달빅 가상 머신을 통해 수행한다. 달빅은 레지스터 기반의 가상머신으로 인터프리터 외에 추적(trace) 기반의 경로 컴파일 정책을 사용하여 달빅 바이트코드를 머신 코드로 번역하여 수행하는 적시 컴파일러를 탑재하였다. 본 논문에서는 달빅 가상 머신의 적시 컴파일러에서 사용하는 추적 단위를 확장하는 장기 추적 방식으로 변경하는 것을 제안한다. 이를 통해 더 많은 바이트코드 명령어를 함께 컴파일하여 컴파일 시간은 길어져도 더 좋은 머신 코드를 생성할 수 있어 하이브리드 컴파일 환경 등을 구축하는 데 유용하다. 실험을 통해 달빅 가상 머신의 성능 향상 가능성을 확인하였다. Google’s mobile platform, Android, executes Java applications by converting Java class files into Dex files and executing them on Dalvik virtual machine. Dalvik is a register-based virtual machine and it includes an interpreter and a just-in-time compiler (JITC) which employs trace-based path compilation for translating Dalvik bytecode to machine code. This paper proposes extending the trace unit for Dalvik JITC so that we can have longer traces. This allows compiling more Dalvik bytecode instructions together to generate better machine code with slightly longer compilation time, which can allow building a hybrid compilation environment. Our experimental results show the possibility of performance improvement for Dalvik virtual machine.

스마트폰 기반 가상 데스크탑과 주변 레거시 디바이스 사이의 연결 중계 소프트웨어

박세진(Sejin Park),박찬익(Chanik Park) 한국정보과학회 2012 정보과학회 컴퓨팅의 실제 논문지 Vol.18 No.1

하나의 물리 머신만으로 복수개의 가상 머신을 지원할 수 있는 가상화 기술이 발달함에 따라, 기업에서는 비용 절감을 위해 가상 데스크탑 기술 도입이 많이 이루어지고 있다, 또한, 일반 사용자들을 위한 클라우드 기반의 가상데스크탑 서비스 역시 많이 소개되고 있는데. 이러한 가상 데스크탑의 효과적인 사용을 위해서는 원격 접속 기술이 필수적이지만, 이는 키보드, 마우스 및 디스플레이가 갖추어져 있는 PC와 같은 정형적인 컴퓨팅 환경에서만 접속이 가능하다. 본 논문에서는 스마트폰을 이용하여, 사용자 주변에서 원격 컴퓨팅 장치로 사용할 수 있는 LCD TV, 키보드 등의 레거시 디바이스를 검색하고 해당 디바이스를 원격지에 있는 가상 데스크탑과 연계시켜, 사용자 주변의 디바이스를 통해 원격 가상 데스크탑 제어를 연계 할 수 있는 중계 소프트웨어에 대한 설계 및 구현 내용을 기술하고 있다. Many enterprises adopt virtual desktop technology to reduce total cost of ownership (TCO) according to the improvement of virtualization technologies that allow one physical machine to run multiple virtual machines. In addition, many cloud based virtual desktop services for common user have been introduced. To support this kind of virtual desktop, remote connection technique is essential. However, current remote connection technique needs a normal computation environment with keyboard, mouse and display. In this paper, we proposed a smartphone based connection broker between remote virtual desktop and local legacy devices. The broker can detect local devices that can be used as a resource of virtual machine such as LCD TV, keyboard and can connect them to remote virtual desktop.