Estudo da efetividade dos mecanismos de compartilhamento de memória em hipervisores

The growing demand for large-scale virtualization environments, such as the ones used in cloud computing, has led to a need for efficient management of computing resources. RAM memory is the one of the most required resources in these environments, and is usually the main factor limiting the number of virtual machines that can run on the physical host. Recently, hypervisors have brought mechanisms for transparent memory sharing between virtual machines in order to reduce the total demand for system memory. These mechanisms “merge” similar pages detected in multiple virtual machines into the same physical memory, using a copy-on-write mechanism in a manner that is transparent to the guest systems. The objective of this study is to present an overview of these mechanisms and also evaluate their performance and effectiveness. The results of two popular hypervisors (VMware and KVM) using different guest operating systems (Linux and Windows) and different workloads (synthetic and real) are presented herein. The results show significant performance differences between hypervisors according to the guest system workloads and execution time.

Event-Code Interaction Directed Test Cases

The Graphical User Interface (GUI) is an integral component of contemporary computer software. A stable and reliable GUI is necessary for correct functioning of software applications. Comprehensive verification of the GUI is a routine part of most software development life-cycles. The input space of a GUI is typically large, making exhaustive verification difficult. GUI defects are often revealed by exercising parts of the GUI that interact with each other. It is challenging for a verification method to drive the GUI into states that might contain defects. In recent years, model-based methods, that target specific GUI interactions, have been developed. These methods create a formal model of the GUI’s input space from specification of the GUI, visible GUI behaviors and static analysis of the GUI’s program-code. GUIs are typically dynamic in nature, whose user-visible state is guided by underlying program-code and dynamic program-state. This research extends existing model-based GUI testing techniques by modelling interactions between the visible GUI of a GUI-based software and its underlying program-code. The new model is able to, efficiently and effectively, test the GUI in ways that were not possible using existing methods. The thesis is this: Long, useful GUI testcases can be created by examining the interactions between the GUI, of a GUI-based application, and its program-code. To explore this thesis, a model-based GUI testing approach is formulated and evaluated. In this approach, program-code level interactions between GUI event handlers will be examined, modelled and deployed for constructing long GUI testcases. These testcases are able to drive the GUI into states that were not possible using existing models. Implementation and evaluation has been conducted using GUITAR, a fully-automated, open-source GUI testing framework.

A Comparison of Geo-tagging in Mobile Internet Browsing Applications on iOS and Android.

Nowadays there is almost no crime committed without a trace of digital evidence, and since the advanced functionality of mobile devices today can be exploited to assist in crime, the need for mobile forensics is imperative. Many of the mobile applications available today, including internet browsers, will request the user’s permission to access their current location when in use. This geolocation data is subsequently stored and managed by that application's underlying database files. If recovered from a device during a forensic investigation, such GPS evidence and track points could hold major evidentiary value for a case. The aim of this paper is to examine and compare to what extent geolocation data is available from the iOS and Android operating systems. We focus particularly on geolocation data recovered from internet browsing applications, comparing the native Safari and Browser apps with Google Chrome, downloaded on to both platforms. All browsers were used over a period of several days at various locations to generate comparable test data for analysis. Results show considerable differences not only in the storage locations and formats, but also in the amount of geolocation data stored by different browsers and on different operating systems.

Reuse and integration of specification logics: the hybridisation perspective

Hybridisation is a systematic process along which the characteristic features of hybrid logic, both at the syntactic and the semantic levels, are developed on top of an arbitrary logic framed as an institution. It also captures the construction of first-order encodings of such hybridised institutions into theories in first-order logic. The method was originally developed to build suitable logics for the specification of reconfigurable software systems on top of whatever logic is used to describe local requirements of each system’s configuration. Hybridisation has, however, a broader scope, providing a fresh example of yet another development in combining and reusing logics driven by a problem from Computer Science. This paper offers an overview of this method, proposes some new extensions, namely the introduction of full quantification leading to the specification of dynamic modalities, and exemplifies its potential through a didactical application. It is discussed how hybridisation can be successfully used in a formal specification course in which students progress from equational to hybrid specifications in a uniform setting, integrating paradigms, combining data and behaviour, and dealing appropriately with systems evolution and reconfiguration.

Hierarchical clustering with multiple-height branch-cut applied to short time-series gene expression data.

Rigid adherence to pre-specified thresholds and static graphical representations can lead to incorrect decisions on merging of clusters. As an alternative to existing automated or semi-automated methods, we developed a visual analytics approach for performing hierarchical clustering analysis of short time-series gene expression data. Dynamic sliders control parameters such as the similarity threshold at which clusters are merged and the level of relative intra-cluster distinctiveness, which can be used to identify "weak-edges" within clusters. An expert user can drill down to further explore the dendrogram and detect nested clusters and outliers. This is done by using the sliders and by pointing and clicking on the representation to cut the branches of the tree in multiple-heights. A prototype of this tool has been developed in collaboration with a small group of biologists for analysing their own datasets. Initial feedback on the tool has been positive.

Dynamically balanced and plausible trajectory planning for human-like characters

International audience

MLCut: exploring multi-level cuts in dendrograms for biological data.

Choosing a single similarity threshold for cutting dendrograms is not sufficient for performing hierarchical clustering analysis of heterogeneous data sets. In addition, alternative automated or semi-automated methods that cut dendrograms in multiple levels make assumptions about the data in hand. In an attempt to help the user to find patterns in the data and resolve ambiguities in cluster assignments, we developed MLCut: a tool that provides visual support for exploring dendrograms of heterogeneous data sets in different levels of detail. The interactive exploration of the dendrogram is coordinated with a representation of the original data, shown as parallel coordinates. The tool supports three analysis steps. Firstly, a single-height similarity threshold can be applied using a dynamic slider to identify the main clusters. Secondly, a distinctiveness threshold can be applied using a second dynamic slider to identify “weak-edges” that indicate heterogeneity within clusters. Thirdly, the user can drill-down to further explore the dendrogram structure - always in relation to the original data - and cut the branches of the tree at multiple levels. Interactive drill-down is supported using mouse events such as hovering, pointing and clicking on elements of the dendrogram. Two prototypes of this tool have been developed in collaboration with a group of biologists for analysing their own data sets. We found that enabling the users to cut the tree at multiple levels, while viewing the effect in the original data, is a promising method for clustering which could lead to scientific discoveries.

Computational Analysis of Intelligent Agents: Social and Strategic Settings

The central motif of this work is prediction and optimization in presence of multiple interacting intelligent agents. We use the phrase `intelligent agents' to imply in some sense, a `bounded rationality', the exact meaning of which varies depending on the setting. Our agents may not be `rational' in the classical game theoretic sense, in that they don't always optimize a global objective. Rather, they rely on heuristics, as is natural for human agents or even software agents operating in the real-world. Within this broad framework we study the problem of influence maximization in social networks where behavior of agents is myopic, but complication stems from the structure of interaction networks. In this setting, we generalize two well-known models and give new algorithms and hardness results for our models. Then we move on to models where the agents reason strategically but are faced with considerable uncertainty. For such games, we give a new solution concept and analyze a real-world game using out techniques. Finally, the richest model we consider is that of Network Cournot Competition which deals with strategic resource allocation in hypergraphs, where agents reason strategically and their interaction is specified indirectly via player's utility functions. For this model, we give the first equilibrium computability results. In all of the above problems, we assume that payoffs for the agents are known. However, for real-world games, getting the payoffs can be quite challenging. To this end, we also study the inverse problem of inferring payoffs, given game history. We propose and evaluate a data analytic framework and we show that it is fast and performant.

Garbage collection optimization for non uniform memory access architectures

Cache-coherent non uniform memory access (ccNUMA) architecture is a standard design pattern for contemporary multicore processors, and future generations of architectures are likely to be NUMA. NUMA architectures create new challenges for managed runtime systems. Memory-intensive applications use the system’s distributed memory banks to allocate data, and the automatic memory manager collects garbage left in these memory banks. The garbage collector may need to access remote memory banks, which entails access latency overhead and potential bandwidth saturation for the interconnection between memory banks. This dissertation makes five significant contributions to garbage collection on NUMA systems, with a case study implementation using the Hotspot Java Virtual Machine. It empirically studies data locality for a Stop-The-World garbage collector when tracing connected objects in NUMA heaps. First, it identifies a locality richness which exists naturally in connected objects that contain a root object and its reachable set— ‘rooted sub-graphs’. Second, this dissertation leverages the locality characteristic of rooted sub-graphs to develop a new NUMA-aware garbage collection mechanism. A garbage collector thread processes a local root and its reachable set, which is likely to have a large number of objects in the same NUMA node. Third, a garbage collector thread steals references from sibling threads that run on the same NUMA node to improve data locality. This research evaluates the new NUMA-aware garbage collector using seven benchmarks of an established real-world DaCapo benchmark suite. In addition, evaluation involves a widely used SPECjbb benchmark and Neo4J graph database Java benchmark, as well as an artificial benchmark. The results of the NUMA-aware garbage collector on a multi-hop NUMA architecture show an average of 15% performance improvement. Furthermore, this performance gain is shown to be as a result of an improved NUMA memory access in a ccNUMA system. Fourth, the existing Hotspot JVM adaptive policy for configuring the number of garbage collection threads is shown to be suboptimal for current NUMA machines. The policy uses outdated assumptions and it generates a constant thread count. In fact, the Hotspot JVM still uses this policy in the production version. This research shows that the optimal number of garbage collection threads is application-specific and configuring the optimal number of garbage collection threads yields better collection throughput than the default policy. Fifth, this dissertation designs and implements a runtime technique, which involves heuristics from dynamic collection behavior to calculate an optimal number of garbage collector threads for each collection cycle. The results show an average of 21% improvements to the garbage collection performance for DaCapo benchmarks.

Joint Optimization for Social Content Delivery in Wireless Networks

Over the last decade, success of social networks has significantly reshaped how people consume information. Recommendation of contents based on user profiles is well-received. However, as users become dominantly mobile, little is done to consider the impacts of the wireless environment, especially the capacity constraints and changing channel. In this dissertation, we investigate a centralized wireless content delivery system, aiming to optimize overall user experience given the capacity constraints of the wireless networks, by deciding what contents to deliver, when and how. We propose a scheduling framework that incorporates content-based reward and deliverability. Our approach utilizes the broadcast nature of wireless communication and social nature of content, by multicasting and precaching. Results indicate this novel joint optimization approach outperforms existing layered systems that separate recommendation and delivery, especially when the wireless network is operating at maximum capacity. Utilizing limited number of transmission modes, we significantly reduce the complexity of the optimization. We also introduce the design of a hybrid system to handle transmissions for both system recommended contents ('push') and active user requests ('pull'). Further, we extend the joint optimization framework to the wireless infrastructure with multiple base stations. The problem becomes much harder in that there are many more system configurations, including but not limited to power allocation and how resources are shared among the base stations ('out-of-band' in which base stations transmit with dedicated spectrum resources, thus no interference; and 'in-band' in which they share the spectrum and need to mitigate interference). We propose a scalable two-phase scheduling framework: 1) each base station obtains delivery decisions and resource allocation individually; 2) the system consolidates the decisions and allocations, reducing redundant transmissions. Additionally, if the social network applications could provide the predictions of how the social contents disseminate, the wireless networks could schedule the transmissions accordingly and significantly improve the dissemination performance by reducing the delivery delay. We propose a novel method utilizing: 1) hybrid systems to handle active disseminating requests; and 2) predictions of dissemination dynamics from the social network applications. This method could mitigate the performance degradation for content dissemination due to wireless delivery delay. Results indicate that our proposed system design is both efficient and easy to implement.

Dynamically managing sensed context data

This thesis reports on an investigation of the feasibility and usefulness of incorporating dynamic management facilities for managing sensed context data in a distributed contextaware mobile application. The investigation focuses on reducing the work required to integrate new sensed context streams in an existing context aware architecture. Current architectures require integration work for new streams and new contexts that are encountered. This means of operation is acceptable for current fixed architectures. However, as systems become more mobile the number of discoverable streams increases. Without the ability to discover and use these new streams the functionality of any given device will be limited to the streams that it knows how to decode. The integration of new streams requires that the sensed context data be understood by the current application. If the new source provides data of a type that an application currently requires then the new source should be connected to the application without any prior knowledge of the new source. If the type is similar and can be converted then this stream too should be appropriated by the application. Such applications are based on portable devices (phones, PDAs) for semi-autonomous services that use data from sensors connected to the devices, plus data exchanged with other such devices and remote servers. Such applications must handle input from a variety of sensors, refining the data locally and managing its communication from the device in volatile and unpredictable network conditions. The choice to focus on locally connected sensory input allows for the introduction of privacy and access controls. This local control can determine how the information is communicated to others. This investigation focuses on the evaluation of three approaches to sensor data management. The first system is characterised by its static management based on the pre-pended metadata. This was the reference system. Developed for a mobile system, the data was processed based on the attached metadata. The code that performed the processing was static. The second system was developed to move away from the static processing and introduce a greater freedom of handling for the data stream, this resulted in a heavy weight approach. The approach focused on pushing the processing of the data into a number of networked nodes rather than the monolithic design of the previous system. By creating a separate communication channel for the metadata it is possible to be more flexible with the amount and type of data transmitted. The final system pulled the benefits of the other systems together. By providing a small management class that would load a separate handler based on the incoming data, Dynamism was maximised whilst maintaining ease of code understanding. The three systems were then compared to highlight their ability to dynamically manage new sensed context. The evaluation took two approaches, the first is a quantitative analysis of the code to understand the complexity of the relative three systems. This was done by evaluating what changes to the system were involved for the new context. The second approach takes a qualitative view of the work required by the software engineer to reconfigure the systems to provide support for a new data stream. The evaluation highlights the various scenarios in which the three systems are most suited. There is always a trade-o↵ in the development of a system. The three approaches highlight this fact. The creation of a statically bound system can be quick to develop but may need to be completely re-written if the requirements move too far. Alternatively a highly dynamic system may be able to cope with new requirements but the developer time to create such a system may be greater than the creation of several simpler systems.

A distributed, compact routing protocol for the Internet

The Internet has grown in size at rapid rates since BGP records began, and continues to do so. This has raised concerns about the scalability of the current BGP routing system, as the routing state at each router in a shortest-path routing protocol will grow at a supra-linearly rate as the network grows. The concerns are that the memory capacity of routers will not be able to keep up with demands, and that the growth of the Internet will become ever more cramped as more and more of the world seeks the benefits of being connected. Compact routing schemes, where the routing state grows only sub-linearly relative to the growth of the network, could solve this problem and ensure that router memory would not be a bottleneck to Internet growth. These schemes trade away shortest-path routing for scalable memory state, by allowing some paths to have a certain amount of bounded “stretch”. The most promising such scheme is Cowen Routing, which can provide scalable, compact routing state for Internet routing, while still providing shortest-path routing to nearly all other nodes, with only slightly stretched paths to a very small subset of the network. Currently, there is no fully distributed form of Cowen Routing that would be practical for the Internet. This dissertation describes a fully distributed and compact protocol for Cowen routing, using the k-core graph decomposition. Previous compact routing work showed the k-core graph decomposition is useful for Cowen Routing on the Internet, but no distributed form existed. This dissertation gives a distributed k-core algorithm optimised to be efficient on dynamic graphs, along with with proofs of its correctness. The performance and efficiency of this distributed k-core algorithm is evaluated on large, Internet AS graphs, with excellent results. This dissertation then goes on to describe a fully distributed and compact Cowen Routing protocol. This protocol being comprised of a landmark selection process for Cowen Routing using the k-core algorithm, with mechanisms to ensure compact state at all times, including at bootstrap; a local cluster routing process, with mechanisms for policy application and control of cluster sizes, ensuring again that state can remain compact at all times; and a landmark routing process is described with a prioritisation mechanism for announcements that ensures compact state at all times.

On the fly type specialization without type analysis

Les langages de programmation typés dynamiquement tels que JavaScript et Python repoussent la vérification de typage jusqu’au moment de l’exécution. Afin d’optimiser la performance de ces langages, les implémentations de machines virtuelles pour langages dynamiques doivent tenter d’éliminer les tests de typage dynamiques redondants. Cela se fait habituellement en utilisant une analyse d’inférence de types. Cependant, les analyses de ce genre sont souvent coûteuses et impliquent des compromis entre le temps de compilation et la précision des résultats obtenus. Ceci a conduit à la conception d’architectures de VM de plus en plus complexes. Nous proposons le versionnement paresseux de blocs de base, une technique de compilation à la volée simple qui élimine efficacement les tests de typage dynamiques redondants sur les chemins d’exécution critiques. Cette nouvelle approche génère paresseusement des versions spécialisées des blocs de base tout en propageant de l’information de typage contextualisée. Notre technique ne nécessite pas l’utilisation d’analyses de programme coûteuses, n’est pas contrainte par les limitations de précision des analyses d’inférence de types traditionnelles et évite la complexité des techniques d’optimisation spéculatives. Trois extensions sont apportées au versionnement de blocs de base afin de lui donner des capacités d’optimisation interprocédurale. Une première extension lui donne la possibilité de joindre des informations de typage aux propriétés des objets et aux variables globales. Puis, la spécialisation de points d’entrée lui permet de passer de l’information de typage des fonctions appellantes aux fonctions appellées. Finalement, la spécialisation des continuations d’appels permet de transmettre le type des valeurs de retour des fonctions appellées aux appellants sans coût dynamique. Nous démontrons empiriquement que ces extensions permettent au versionnement de blocs de base d’éliminer plus de tests de typage dynamiques que toute analyse d’inférence de typage statique.

Efficient Image and Video Representations for Retrieval

Image (Video) retrieval is an interesting problem of retrieving images (videos) similar to the query. Images (Videos) are represented in an input (feature) space and similar images (videos) are obtained by finding nearest neighbors in the input representation space. Numerous input representations both in real valued and binary space have been proposed for conducting faster retrieval. In this thesis, we present techniques that obtain improved input representations for retrieval in both supervised and unsupervised settings for images and videos. Supervised retrieval is a well known problem of retrieving same class images of the query. We address the practical aspects of achieving faster retrieval with binary codes as input representations for the supervised setting in the first part, where binary codes are used as addresses into hash tables. In practice, using binary codes as addresses does not guarantee fast retrieval, as similar images are not mapped to the same binary code (address). We address this problem by presenting an efficient supervised hashing (binary encoding) method that aims to explicitly map all the images of the same class ideally to a unique binary code. We refer to the binary codes of the images as `Semantic Binary Codes' and the unique code for all same class images as `Class Binary Code'. We also propose a new class­ based Hamming metric that dramatically reduces the retrieval times for larger databases, where only hamming distance is computed to the class binary codes. We also propose a Deep semantic binary code model, by replacing the output layer of a popular convolutional Neural Network (AlexNet) with the class binary codes and show that the hashing functions learned in this way outperforms the state­ of ­the art, and at the same time provide fast retrieval times. In the second part, we also address the problem of supervised retrieval by taking into account the relationship between classes. For a given query image, we want to retrieve images that preserve the relative order i.e. we want to retrieve all same class images first and then, the related classes images before different class images. We learn such relationship aware binary codes by minimizing the similarity between inner product of the binary codes and the similarity between the classes. We calculate the similarity between classes using output embedding vectors, which are vector representations of classes. Our method deviates from the other supervised binary encoding schemes as it is the first to use output embeddings for learning hashing functions. We also introduce new performance metrics that take into account the related class retrieval results and show significant gains over the state­ of­ the art. High Dimensional descriptors like Fisher Vectors or Vector of Locally Aggregated Descriptors have shown to improve the performance of many computer vision applications including retrieval. In the third part, we will discuss an unsupervised technique for compressing high dimensional vectors into high dimensional binary codes, to reduce storage complexity. In this approach, we deviate from adopting traditional hyperplane hashing functions and instead learn hyperspherical hashing functions. The proposed method overcomes the computational challenges of directly applying the spherical hashing algorithm that is intractable for compressing high dimensional vectors. A practical hierarchical model that utilizes divide and conquer techniques using the Random Select and Adjust (RSA) procedure to compress such high dimensional vectors is presented. We show that our proposed high dimensional binary codes outperform the binary codes obtained using traditional hyperplane methods for higher compression ratios. In the last part of the thesis, we propose a retrieval based solution to the Zero shot event classification problem - a setting where no training videos are available for the event. To do this, we learn a generic set of concept detectors and represent both videos and query events in the concept space. We then compute similarity between the query event and the video in the concept space and videos similar to the query event are classified as the videos belonging to the event. We show that we significantly boost the performance using concept features from other modalities.

SEQUENTIAL DECISIONS AND PREDICTIONS IN NATURAL LANGUAGE PROCESSING

Natural language processing has achieved great success in a wide range of ap- plications, producing both commercial language services and open-source language tools. However, most methods take a static or batch approach, assuming that the model has all information it needs and makes a one-time prediction. In this disser- tation, we study dynamic problems where the input comes in a sequence instead of all at once, and the output must be produced while the input is arriving. In these problems, predictions are often made based only on partial information. We see this dynamic setting in many real-time, interactive applications. These problems usually involve a trade-off between the amount of input received (cost) and the quality of the output prediction (accuracy). Therefore, the evaluation considers both objectives (e.g., plotting a Pareto curve). Our goal is to develop a formal understanding of sequential prediction and decision-making problems in natural language processing and to propose efficient solutions. Toward this end, we present meta-algorithms that take an existent batch model and produce a dynamic model to handle sequential inputs and outputs. Webuild our framework upon theories of Markov Decision Process (MDP), which allows learning to trade off competing objectives in a principled way. The main machine learning techniques we use are from imitation learning and reinforcement learning, and we advance current techniques to tackle problems arising in our settings. We evaluate our algorithm on a variety of applications, including dependency parsing, machine translation, and question answering. We show that our approach achieves a better cost-accuracy trade-off than the batch approach and heuristic-based decision- making approaches. We first propose a general framework for cost-sensitive prediction, where dif- ferent parts of the input come at different costs. We formulate a decision-making process that selects pieces of the input sequentially, and the selection is adaptive to each instance. Our approach is evaluated on both standard classification tasks and a structured prediction task (dependency parsing). We show that it achieves similar prediction quality to methods that use all input, while inducing a much smaller cost. Next, we extend the framework to problems where the input is revealed incremen- tally in a fixed order. We study two applications: simultaneous machine translation and quiz bowl (incremental text classification). We discuss challenges in this set- ting and show that adding domain knowledge eases the decision-making problem. A central theme throughout the chapters is an MDP formulation of a challenging problem with sequential input/output and trade-off decisions, accompanied by a learning algorithm that solves the MDP.