3,854 research outputs found
Secure in-VM monitoring using hardware virtualization
No full textKernel-level attacks or rootkits can compromise the security of an operating system by executing with the privilege of the kernel. Current approaches use virtualization to gain higher privilege over these attacks, and isolate security tools from the untrusted guest VM by moving them out and placing them in a separate trusted VM. Although out-of-VM isolation can help ensure security, the added overhead of world-switches between the guest VMs for each invocation of the monitor makes this approach unsuitable for many applications, especially fine-grained monitoring. In this paper, we present Secure In-VM Monitoring (SIM), a general-purpose framework that enables security monitoring applications to be placed back in the untrusted guest VM for efficiency without sacrificing the security guarantees provided by running them outside of the VM. We utilize contemporary hardware memory protection and hardware virtualization features available in recent processors to create a hypervisor protected address space where a monitor can execute and access data in native speeds and to which execution is transferred in a controlled manner that does not require hypervisor involvement. We have developed a prototype into KVM utilizing Intel VT hardware virtualization technology. We have also developed two representative applications for the Windows OS that monitor system calls and process creations. Our microbenchmarks show at least 10 times performance improvement in invocation of a monitor inside SIM over a monitor residing in another trusted VM. With a systematic security analysis of SIM against a number of possible threats, we show that SIM provides at least the same security guarantees as what can be achieved by out-of-VM monitors
Dynamix on the Frame VM: Declarative dynamic semantics on a VM using scopes as frames
No full textOver the years virtual machines (VMs) have been created to abstract over computer hardware. This simplified code generation and allowed for easy portability between hardware platforms. These VMs are however highly tailored to a particular runtime model. This improves the execution speed, but places restrictions on the types of languages that the VM supports. In this thesis the Frame VM was developed as a VM that supports many different types of languages in a principled way. Achieving this is done by basing the VM on language independent models of memory and control flow. Usage of the scopes-as-frames paradigm and control frames resulted in an instruction set that is relatively small at its core, but does allow for the construction of complex control flow. As an effect, many different programming languages can be compiled to the Frame VM. In addition to this VM, a Domain Specific Language (DSL) for executable semantics of programming languages was created. This language, Dynamix, allows for a modular approach to writing the semantics of a language. Additionally, Dynamix provides a meta-compiler that uses these semantics of a language to compile programs to the Frame VM. To validate the Frame VM, direct compilers for Rust and Prolog have been created in a student project and compilers for Scheme and Tiger were created using Dynamix. Using these semantics of Scheme and Tiger, it was possible to execute programs containing usage of call/cc and a suite of Tiger benchmark programs. Furthermore, the control flow of Tiger was extended with exceptions and generator functions. This extension did not require any changes to the existing semantics, showing the modularity of control achieved when using Dynamix and the Frame VM.Computer Scienc
Locality-aware dynamic VM reconfiguration on MapReduce clouds
No full textCloud computing based on system virtualization, has been expanding its services to distributed data-intensive platforms such as MapReduce and Hadoop. Such a distributed platform on clouds runs in a virtual cluster consisting of a number of virtual machines. In the virtual cluster, demands on computing resources for each node may fluctuate, due to data locality and task behavior. However, current cloud services use a static cluster configuration, fixing or manually adjusting the computing capability of each virtual machine (VM). The fixed homogeneous VM configuration may not adapt to changing resource demands in individual nodes.
In this paper, we propose a dynamic VM reconfiguration technique for data-intensive computing on clouds, called Dynamic Resource Reconfiguration (DRR). DRR can adjust the computing capability of individual VMs to maximize the utilization of resources. Among several factors causing resource imbalance in the Hadoop platforms, this paper focuses on data locality. Although assigning tasks on the nodes containing their input data can improve the overall performance of a job significantly, the fixed computing capability of each node may not allow such locality-aware scheduling. DRR dynamically increases or decreases the computing capability of each node to enhance locality-aware task scheduling. We evaluate the potential performance improvement of DRR on a 100-node cluster, and its detailed behavior on a small scale cluster with constrained network bandwidth. On the 100-node cluster, DRR can improve the throughput of Hadoop jobs by 15% on average, and 41% on the private cluster with the constrained network connection
Coupling between flexible ship and liquid sloshing using potential flow analysis
No full textThe significant increase in demand for Liquefied Natural Gas (LNG) and the economic aspects of its transportation resulted in increases in the number and size of LNG carriers. One of the design issues for LNG carriers is the sloshing phenomenon because containment systems widely used nowadays have no internal structures. Furthermore, because the weights of ship and cargo are comparable and ship operators want more flexible operations allowing partial fillings in tanks, the coupling effect between ship motions and sloshing requires further investigation, including the effect of ship distortion.The previous study on coupling between rigid body and sloshing shows good agreement between methods of prediction and measurements[1,2]. Hence, in this paper the potential flow approach adopted for the coupling effect between rigid body ship motion and sloshing is extended to flexible ship-partially filled tank system, using the de-singularised Rankine source method. In this case, the global deflection of the flexible ship is used for application of the body boundary condition on the partially filled tank. The aim of this paper is to investigate the influence of hull flexibility on the hydrodynamic forces and moments associated with liquid sloshing and vice versa, as well as the dynamic characteristics (e.g. resonance frequencies) of the whole system. As there are no experimental results available, the method is validated by comparing hydrodynamic forces from sloshing obtained using rigid and flexible body approaches. The coupling effect between flexible ship and sloshing in partially filled tanks is investigated for an idealized LNG Carrier in beam regular waves, considering different partial filling scenarios
웨이퍼 가장자리 식각 조절을 위한 PI-VM 진단기술 개발
No full text학위논문(석사) -- 서울대학교대학원 : 공과대학 에너지시스템공학부, 2024. 2. 김곤호.The semiconductor industry aims to enhance manufacturing efficiency as a primary goal to increase profits. There are three main methods to achieve this. First, by increasing the number of layers or reducing the pitch size of cells to improve integration density. Second, by ensuring process uniformity across the entire wafer to increase die yield. Third, by maximizing production speed. As a strategy to maximize productivity, virtual metrology (VM) has been proposed, which predicts equipment and process performance using data generated from process equipment. Since most unit processes utilize plasma, recent research suggests that it is advantageous to include plasma information (PI) parameters when applying VM to plasma processes. PI literally refers to information about the process plasma within the chamber and typically consists of non-linearly combined equipment and sensor data. Therefore, PI-VM, which includes these PI parameters in the input dataset, can mediate the complex non-linearity of plasma in linear VM models and has high predictive power in plasma processes. However, traditional VM, developed to predict wafer-level process results based on zero-dimensional data, i.e., data averaged over the entire wafer, lacks spatial resolution and therefore has limitations in identifying key features affecting process uniformity within the wafer, which corresponds to a one-dimensional concept. The need for spatial resolution in PI-VM for ensuring plasma process uniformity is due to the plasma's distribution within the chamber caused by various physical phenomena. Such distributions of plasma characteristics affect process uniformity. Particularly in semiconductor etching processes, the plasma density is higher at the center of the chamber and decreases as it moves outward, leading to uniformity issues such as reduced etching rate and tilting at the wafer edge region. To solve these problems, manufacturing sites adjust recipe parameters or utilize equipment parts. In etching processes using capacitively coupled plasma (CCP), a part known as edge ring or focus ring near the wafer edge region is used to adjust the sheath at the wafer edge region by altering the height of the edge ring or changing its RF characteristics. Therefore, to effectively utilize edge ring for process control, it is necessary to develop a methodology that systematizes its role and usage. For this purpose, the concept of plasma distribution was introduced into PI-VM. By operating the powered edge ring (PER) and observing the changes in plasma characteristics, PI parameters were derived, and areas where these changes occur were distinguished to develop PI-VM for each region, referred to as Regional PI-VM. In this study, diagnostic equipment was installed in very high frequency (VHF) CCP etching equipment, and plasma distribution data were acquired while adjusting the RF frequency and power applied to the PER. When additional power was applied to the PER, the plasma characteristics were locally enhanced near the PER. Based on this understanding of plasma phenomena, the wafer space was divided, and regional PI-VMs were developed to predict the etching uniformity on a 300 mm wafer for each region. The results of the regional PI-VM showed higher predictive accuracy in all regions compared to the conventional global PI-VM. PER at 2 MHz altered the ion energy flux, which significantly contributes to mask etching in the wafer edge region. PER at 400 kHz hardly caused any changes in the ion energy flux and etching rate within the wafer space. Through the analysis of selected key features for each region, the effects of PER operating parameters on plasma characteristics and their impact on process control in the wafer edge region could be predicted. The development of regional PI-VM is an important proactive research for the development of process recipes and equipment parts for process uniformity control, offering the possibility to alleviate process issues caused by high density in the center. As this methodology advances, it is expected to contribute to the selection of control parameters for spatial controllers that can selectively control the regions necessary to ensure process uniformity.반도체 산업은 수익 증대를 위해 제조 효율을 향상시키는 것을 주요 목표로 하고 있다. 이를 달성하기 위한 방법은 크게 3가지로 나눌 수 있다. 첫째, 적층 단수를 증가시키거나 셀의 피치 크기를 줄여 집적도를 향상시키는 것이다. 둘째, 웨이퍼 전체에 걸쳐 공정 균일도를 확보하여 다이 수율을 높이는 것이다. 셋째, 생산 속도를 최대화하는 것이다. 특히, 생산성을 극대화하기 위한 전략으로 공정 장비에서 생산되는 데이터를 활용하여 장비 및 공정 성능을 예측하는 가상 계측(VM) 방법론이 제안되었다. 최근의 반도체 공정은 대부분의 단위 공정에서 플라즈마를 활용하고 있기 때문에, 최신 연구 결과에 따르면 플라즈마 공정에 VM을 적용할 때 플라즈마 정보(PI) 인자를 포함시키는 것이 유리하다. PI는 글자 그대로 챔버 내 공정 플라즈마에 대한 정보를 가리키며, 일반적으로 장비 및 센서 데이터가 비선형적으로 조합된 형태이다. 따라서 이러한 PI 인자를 입력 데이터 세트에 포함시킨 PI-VM은 선형 VM 모델에서도 플라즈마의 복잡한 비선형성을 중재할 수 있으며 플라즈마 공정에서 높은 예측력을 갖는다. 그러나 기존의 VM은 웨이퍼 전체를 평균한 데이터, 즉 공간 분해능은 가지고 있지 않은 0차원의 데이터를 바탕으로 웨이퍼 단위의 공정 결과를 예측하도록 개발되기 때문에, 1차원 개념에 해당하는 웨이퍼 내 공정 균일도에 영향을 미치는 주요 인자를 파악하는 데는 한계가 있다. 플라즈마 공정 균일도 확보를 위한 PI-VM에 공간 분해능이 필요한 이유는 플라즈마가 다양한 물리적 현상들에 의해 챔버 내에서 분포를 이루기 때문이다. 이러한 플라즈마의 특성 분포는 공정 균일도에도 영향을 미친다. 특히 반도체 식각 공정에서는 챔버 중심부의 플라즈마 밀도가 높고 중심에서 멀어질수록 밀도가 감소하는 경향이 있어, 이는 웨이퍼 가장자리에서의 식각 속도 감소, 틸팅 등의 균일도 문제를 야기한다. 이러한 문제를 해결하기 위해, 제조 현장에서는 레시피 파라미터들을 조절하거나 설비 부품을 활용한다. 용량 결합 플라즈마(CCP)를 이용하는 식각 공정에서는 웨이퍼 가장자리 부근의 에지 링 또는 포커스 링이라 불리는 부품을 이용한다. 에지 링의 높이를 조절하거나 RF 특성을 변화시켜서 웨이퍼 가장자리의 쉬스를 조절한다. 따라서 에지 링을 공정 제어에 효과적으로 활용하기 위해서는 그 역할과 사용 방법을 체계화하는 방법론 개발이 필요하다. 이를 위해 PI-VM에 플라즈마 분포 개념을 도입함으로써, 전력 인가 에지 링(PER)을 운용함에 따라 발생하는 플라즈마 특성의 변화를 확인하여 PI 파라미터들을 도출하고 이러한 변화가 발생하는 영역을 구분 지어, 영역별로 PI-VM을 개발하는 방법을 제시한다. 이를 영역별 PI VM(Regional PI-VM)이라 한다. 본 연구에서는 초고주파(VHF) CCP 식각 장비에 진단계를 설치하고, PER에 인가하는 RF 주파수와 전력을 조절하며 플라즈마 분포 데이터를 취득하였다. VHF에 의해 생성된 플라즈마에 더해 별도의 PER 전력을 인가했을 때 PER 근방에서 국부적으로 플라즈마 특성이 강화되었다. 이러한 플라즈마 현상학을 기반으로 웨이퍼 공간을 나누고, 300 mm 웨이퍼 상의 식각 균일도를 예측하기 위한 regional PI-VM을 개발하였다. Regional PI-VM 결과는 기존의 공간 평균 PI-VM (Global PI-VM) 결과보다 모든 영역에서 더 높은 예측력을 보였다. PER 2 MHz는 웨이퍼 가장자리 영역에서 마스크 식각에 주요하게 기여하는 이온 에너지 플럭스를 변화시켰다. PER 400 kHz는 웨이퍼 공간 내 이온 에너지 플럭스 및
식각률 변화를 거의 일으키지 않았다. 영역 별로 선택된 주요 인자에 대한 분석을 통해 PER 운전 인자가 플라즈마 특성에 미치는 영향을 파악하고 웨이퍼 가장자리 영역의 공정 제어에 미치는 영향을 예측할 수 있었다. Regional PI-VM 개발은 공정 균일도 제어를 위한 공정 레시피 개발 및 설비 부품 개발에 중요한 선제적 연구로서, 중심부 고밀도에 의한 공정 문제들도 완화할 수 있는 가능성을 제시한다. 이 방법론이 발전하면, 공정 균일도 확보에 필요한 제어 영역을 찾아 선택적으로 제어할 수 있는 공간 제어기의 제어 인자를 선정하는 데 기여할 것으로 기대한다.Abstract ⅰ
Chapter 1. Introduction 1
1.1. Plasma Information based Virtual Metrology 1
1.2. Plasma Distribution of VHF-driven CCP 3
1.3. Regional PI-VM 5
Chapter 2. Experimental Setup 7
2.1. Equipment Details 7
2.2. Process Details 9
2.3. Coupon Sample Processing and Data Acquistion 12
Chapter 3. PI paramters and Regions for Powered Edge Ring 15
3.1. Plasma Characteristics by Powered Edge Ring 15
3.2. New Region formed by PER operation 17
Chapter 4. PI-VM results for Etch Process 22
4.1. PI-VMs for PER with 2 MHz : High Density Mode 22
4.2. PI-VMs for PER with 400 kHz : Low Density Mode 28
Chapter 5. Roles of Regional PI-VM 34
5.1. Feature Analysis : Ion Energy Flux and Sheath Potential 34
5.2. PER Operating Parameters : Frequency, Power 37
Chapter 6. Conclusion 39
Bibliography 42
초 록 45석
A poor women's pedagogy' : an exploration of learning in a housing social movement
No full textIncludes bibliographical references (p. 270-284).This study examines the critical role that adult education played in a housing social movement whose membership was mainly poor African women in informal settlements. In this social movement women have combined learning with the struggle to obtain social goods from the state. The study explores the interconnectedness between learning, development and social change. The conceptual framework developed from a feminist critique of popular education was applied in the methodology and yielded insights with regard to the learning of VM women. The feminist critique allowed for an exploration of the contradictions within popular education and people-centred development. In addition it provided a vocabulary to explain the learning and agency of VM women. The conceptual framework allowed me to argue that learning is contextual, and to analyse and understand learning in the micro-context (VM and the life changes and learning of VM women) it is necessary to examine the interaction between the macro-- (political, economic and policy context of South Africa) and micro-contexts. The interaction of these contexts has brought political opportunities to mobilise the agency of poor African women who were seeking solutions to their housing problems
Hydor-impact, fluid-structure interaction and structural response of modern racing yacht
No full textIn recent years, faster, lighter and bigger are the key issues in a modern racing yachtfor extreme performance. As a result, many yachts have experienced various structuralfailures caused by the hydrodynamic impact or ’hydro-impact’ phenomenon byslamming.The structural failure by hydro-impact originates from the facts that the externalhydro-impact load and fluid structure interaction effect is somewhat misled and whenapplying the load into current structural design, the ’dynamic’ load is typically, manipulatedin a ’static’ way with fluid structure interaction effect, generally, ignored.In this thesis, the hydro-impact load by slamming, its fluid structure interaction effectand dynamic response of the local structure of the yacht are studied.Firstly, to acquire insight into the hydro-impact phenomenon, a series of drop testsand seakeeping-slamming tests are carried out with various sensing instruments ofpressure transducers, accelerometer and ’slam patch system’ - a specific applicationform of generally known pressure panel - are installed. The slam patch system isdesigned and implemented to investigate the hydro-impact loads and fluid structureinteraction effect of slamming. Afterward, the measured hydro-impact loads are summarisedvia statistical manipulations with regard to pressure and duration time.Secondly, impact pressure by the rules and regulations of various organisations areprovided to compare it with the experimental results and structural response calculations.The applicability of the rules and regulations on the high performance racingyacht is also pointed out.Finally, the manipulated loads are used as input data to simulate the transientresponse of local structure of the yacht structure.Throughout this study, the dynamic and fluid structure interaction effect by hydroimpactphenomenon on local composite structure can be easily visualised and calculatedin a conservative way through conventional finite element analysis work
Speeding Up VM Image Distribution for Cloud Data Centers
No full textAn efficient scheme of virtual machine image dissemination, which takes up a major portion of VM provisioning time,
is key to ensuring that the service can be provisioned and delivered to clients in a timely manner. This paper
presents a novel peer-to-peer protocol to efficiently distribute virtual machine images in a cloud data center.
The primary idea of our proposal is to expedite the peer-to-peer content delivery by taking advantage of a high
level of similarity that oftentimes exists among different VM images. The efficacy of our protocol is validated
through an evaluation that demonstrates substantial performance gains over existing approaches
Highly VM-Scalable SSD in Cloud Storage Systems
No full textSolid-state drives (SSDs) are widely used in cloud storage. As the capacity of an SSD has been increasing, it has become common for many virtual machines (VMs) to share a single SSD to maximize resource utilization. However, this sharing can degrade the efficiency of internal operations, such as garbage collection, resulting in increased latencies. Existing literature in this field has mostly focused on interdevice isolation considering the storage device as a black-box entity or presumed an SSD to be shared by up to only eight VMs. In this study, we first analyze a realistic SSD usage environment in cloud systems and identify that block-level data isolation (BDI) should be guaranteed to efficiently scale up the number of VMs in an SSD with minimum latency increases. However, previous schemes cannot work efficiently with BDI when the SSD is shared by dozens of VMs. Based on this analysis, we propose an SSD internal resource management scheme in a cloud environment, called highly VM-scalable SSD (VMS). VMS dynamically partitions physical resources and allocates them to VMs, while the VMs share global buffer blocks to lower latency during abrupt fluctuations of write I/O intensities. Our experimental results show up to 29% of latency reduction. VMS exhibits reduced latencies even in the experiment with 64 VMs, where existing schemes do not function normally.
- …
