Orquestração de migração massiva de máquinas virtuais baseada em análise cíclica para ambientes de computação na nuvem.

Um das principais características da tecnologia de virtualização é a Live Migration, que permite que máquinas virtuais sejam movimentadas entre máquinas físicas sem a interrupção da execução. Esta característica habilita a implementação de políticas mais sofisticadas dentro de um ambiente de computação na nuvem, como a otimização de uso de energia elétrica e recursos computacionais. Entretanto, a Live Migration pode impor severa degradação de desempenho nas aplicações das máquinas virtuais e causar diversos impactos na infraestrutura dos provedores de serviço, como congestionamento de rede e máquinas virtuais co-existentes nas máquinas físicas. Diferente de diversos estudos, este estudo considera a carga de trabalho da máquina virtual um importante fator e argumenta que escolhendo o momento adequado para a migração da máquina virtual pode-se reduzir as penalidades impostas pela Live Migration. Este trabalho introduz a Application-aware Live Migration (ALMA), que intercepta as submissões de Live Migration e, baseado na carga de trabalho da aplicação, adia a migração para um momento mais favorável. Os experimentos conduzidos neste trabalho mostraram que a arquitetura reduziu em até 74% o tempo das migrações para os experimentos com benchmarks e em até 67% os experimentos com carga de trabalho real. A transferência de dados causada pela Live Migration foi reduzida em até 62%. Além disso, o presente introduz um modelo que faz a predição do custo da Live Migration para a carga de trabalho e também um algoritmo de migração que não é sensível à utilização de memória da máquina virtual.

Design and implementation of a windows-based parallel computing environment for large scale optimization

A parallel computing environment to support optimization of large-scale engineering systems is designed and implemented on Windows-based personal computer networks, using the master-worker model and the Parallel Virtual Machine (PVM). It is involved in decomposition of a large engineering system into a number of smaller subsystems optimized in parallel on worker nodes and coordination of subsystem optimization results on the master node. The environment consists of six functional modules, i.e. the master control, the optimization model generator, the optimizer, the data manager, the monitor, and the post processor. Object-oriented design of these modules is presented. The environment supports steps from the generation of optimization models to the solution and the visualization on networks of computers. User-friendly graphical interfaces make it easy to define the problem, and monitor and steer the optimization process. It has been verified by an example of a large space truss optimization. (C) 2004 Elsevier Ltd. All rights reserved.

Dynamic binary translation using run-time feedbacks

Dynamic binary translation is the process of translating, modifying and rewriting executable (binary) code from one machine to another at run-time. This process of low-level re-engineering consists of a reverse engineering phase followed by a forward engineering phase. UQDBT, the University of Queensland Dynamic Binary Translator, is a machine-adaptable translator. Adaptability is provided through the specification of properties of machines and their instruction sets, allowing the support of different pairs of source and target machines. Most binary translators are closely bound to a pair of machines, making analyses and code hard to reuse. Like most virtual machines, UQDBT performs generic optimizations that apply to a variety of machines. Frequently executed code is translated to native code by the use of edge weight instrumentation, which makes UQDBT converge more quickly than systems based on instruction speculation. In this paper, we describe the architecture and run-time feedback optimizations performed by the UQDBT system, and provide results obtained in the x86 and SPARC® platforms.

Canonical solution groups of a homomorphism : manipulator kinematics, nonparametric regression and distributed object systems

The kinematic mapping of a rigid open-link manipulator is a homomorphism between Lie groups. The homomorphisrn has solution groups that act on an inverse kinematic solution element. A canonical representation of solution group operators that act on a solution element of three and seven degree-of-freedom (do!) dextrous manipulators is determined by geometric analysis. Seven canonical solution groups are determined for the seven do! Robotics Research K-1207 and Hollerbach arms. The solution element of a dextrous manipulator is a collection of trivial fibre bundles with solution fibres homotopic to the Torus. If fibre solutions are parameterised by a scalar, a direct inverse funct.ion that maps the scalar and Cartesian base space coordinates to solution element fibre coordinates may be defined. A direct inverse pararneterisation of a solution element may be approximated by a local linear map generated by an inverse augmented Jacobian correction of a linear interpolation. The action of canonical solution group operators on a local linear approximation of the solution element of inverse kinematics of dextrous manipulators generates cyclical solutions. The solution representation is proposed as a model of inverse kinematic transformations in primate nervous systems. Simultaneous calibration of a composition of stereo-camera and manipulator kinematic models is under-determined by equi-output parameter groups in the composition of stereo-camera and Denavit Hartenberg (DH) rnodels. An error measure for simultaneous calibration of a composition of models is derived and parameter subsets with no equi-output groups are determined by numerical experiments to simultaneously calibrate the composition of homogeneous or pan-tilt stereo-camera with DH models. For acceleration of exact Newton second-order re-calibration of DH parameters after a sequential calibration of stereo-camera and DH parameters, an optimal numerical evaluation of DH matrix first order and second order error derivatives with respect to a re-calibration error function is derived, implemented and tested. A distributed object environment for point and click image-based tele-command of manipulators and stereo-cameras is specified and implemented that supports rapid prototyping of numerical experiments in distributed system control. The environment is validated by a hierarchical k-fold cross validated calibration to Cartesian space of a radial basis function regression correction of an affine stereo model. Basic design and performance requirements are defined for scalable virtual micro-kernels that broker inter-Java-virtual-machine remote method invocations between components of secure manageable fault-tolerant open distributed agile Total Quality Managed ISO 9000+ conformant Just in Time manufacturing systems.

Modeling communication services for rapid realization

The availability and pervasiveness of new communication services, such as mobile networks and multimedia communication over digital networks, has resulted in strong demands for approaches to modeling and realizing customized communication systems. The stovepipe approach used to develop today's communication applications is no longer effective because it results in a lengthy and expensive development cycle. To address this need, the Communication Virtual Machine (CVM) technology has been developed by researchers at Florida International University. The CVM technology includes the Communication Modeling Language (CML) and the platform, CVM, to model and rapidly realize communication models. ^ In this dissertation, we investigate the basic communication primitives needed to capture and specify an end-user's requirements for communication-intensive applications, and how these specifications can be automatically realized. To identify the basic communication primitives, we perform a feature analysis on a set of communication-intensive scenarios from the healthcare domain. Based on the feature analysis, we define a new version of CML that includes the meta-model definition (abstract syntax and static semantics) and a partial behavior model (operational semantics). To validate our CML definition, we present a case study that shows how one of the scenarios from the healthcare domain is modeled and automatically realized. ^

Abstractions to support dynamic adaptation of communication frameworks for user-centric communication

The convergence of data, audio and video on IP networks is changing the way individuals, groups and organizations communicate. This diversity of communication media presents opportunities for creating synergistic collaborative communications. This form of collaborative communication is however not without its challenges. The increasing number of communication service providers coupled with a combinatorial mix of offered services, varying Quality-of-Service and oscillating pricing of services increases the complexity for the user to manage and maintain ‘always best’ priced or performance services. Consumers have to manually manage and adapt their communication in line with differences in services across devices, networks and media while ensuring that the usage remain consistent with their intended goals. This dissertation proposes a novel user-centric approach to address this problem. The proposed approach aims to reduce the aforementioned complexity to the user by (1) providing high-level abstractions and a policy based methodology for automated selection of the communication services guided by high-level user policies and (2) providing services through the seamless integration of multiple communication service providers and providing an extensible framework to support the integration of multiple communication service providers. The approach was implemented in the Communication Virtual Machine (CVM), a model-driven technology for realizing communication applications. The CVM includes the Network Communication Broker, the layer responsible for providing a network-independent API to the upper layers of CVM. The initial prototype for the NCB supported only a single communication framework which limited the number, quality and types of services available. Experimental evaluation of the approach show the additional overhead of the approach is minimal compared to the individual communication services frameworks. Additionally the automated approach proposed out performed the individual communication services frameworks for cross framework switching.

A domain specific modeling approach for coordinating user-centric communication services

Rapid advances in electronic communication devices and technologies have resulted in a shift in the way communication applications are being developed. These new development strategies provide abstract views of the underlying communication technologies and lead to the so-called user-centric communication applications. One user-centric communication (UCC) initiative is the Communication Virtual Machine (CVM) technology, which uses the Communication Modeling Language (CML) for modeling communication services and the CVM for realizing these services. In communication-intensive domains such as telemedicine and disaster management, there is an increasing need for user-centric communication applications that are domain-specific and that support the dynamic coordination of communication services commonly found in collaborative communication scenarios. However, UCC approaches like the CVM offer little support for the dynamic coordination of communication services resulting from inherent dependencies between individual steps of a collaboration task. Users either have to manually coordinate communication services, or reply on a process modeling technique to build customized solutions for services in a specific domain that are usually costly, rigidly defined and technology specific. ^ This dissertation proposes a domain-specific modeling approach to address this problem by extending the CVM technology with communication-specific abstractions of workflow concepts commonly found in business processes. The extension involves (1) the definition of the Workflow Communication Modeling Language (WF-CML), a superset of CML, and (2) the extension of the functionality of CVM to process communication-specific workflows. The definition of WF-CML includes the meta-model and the dynamic semantics for control constructs and concurrency. We also extended the CVM prototype to handle the modeling and realization of WF-CML models. A comparative study of the proposed approach with other workflow environments validates the claimed benefits of WF-CML and CVM.^

A generic model of execution for synthesizing domain-specific models

Software engineering researchers are challenged to provide increasingly more powerful levels of abstractions to address the rising complexity inherent in software solutions. One new development paradigm that places models as abstraction at the forefront of the development process is Model-Driven Software Development (MDSD). MDSD considers models as first class artifacts, extending the capability for engineers to use concepts from the problem domain of discourse to specify apropos solutions. A key component in MDSD is domain-specific modeling languages (DSMLs) which are languages with focused expressiveness, targeting a specific taxonomy of problems. The de facto approach used is to first transform DSML models to an intermediate artifact in a HLL e.g., Java or C++, then execute that resulting code.^ Our research group has developed a class of DSMLs, referred to as interpreted DSMLs (i-DSMLs), where models are directly interpreted by a specialized execution engine with semantics based on model changes at runtime. This execution engine uses a layered architecture and is referred to as a domain-specific virtual machine (DSVM). As the domain-specific model being executed descends the layers of the DSVM the semantic gap between the user-defined model and the services being provided by the underlying infrastructure is closed. The focus of this research is the synthesis engine, the layer in the DSVM which transforms i-DSML models into executable scripts for the next lower layer to process.^ The appeal of an i-DSML is constrained as it possesses unique semantics contained within the DSVM. Existing DSVMs for i-DSMLs exhibit tight coupling between the implicit model of execution and the semantics of the domain, making it difficult to develop DSVMs for new i-DSMLs without a significant investment in resources.^ At the onset of this research only one i-DSML had been created for the user- centric communication domain using the aforementioned approach. This i-DSML is the Communication Modeling Language (CML) and its DSVM is the Communication Virtual machine (CVM). A major problem with the CVM's synthesis engine is that the domain-specific knowledge (DSK) and the model of execution (MoE) are tightly interwoven consequently subsequent DSVMs would need to be developed from inception with no reuse of expertise.^ This dissertation investigates how to decouple the DSK from the MoE and subsequently producing a generic model of execution (GMoE) from the remaining application logic. This GMoE can be reused to instantiate synthesis engines for DSVMs in other domains. The generalized approach to developing the model synthesis component of i-DSML interpreters utilizes a reusable framework loosely coupled to DSK as swappable framework extensions.^ This approach involves first creating an i-DSML and its DSVM for a second do- main, demand-side smartgrid, or microgrid energy management, and designing the synthesis engine so that the DSK and MoE are easily decoupled. To validate the utility of the approach, the SEs are instantiated using the GMoE and DSKs of the two aforementioned domains and an empirical study to support our claim of reduced developmental effort is performed.^

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.

A Generic Model of Execution for Synthesizing Domain-Specific Models

Software engineering researchers are challenged to provide increasingly more pow- erful levels of abstractions to address the rising complexity inherent in software solu- tions. One new development paradigm that places models as abstraction at the fore- front of the development process is Model-Driven Software Development (MDSD). MDSD considers models as first class artifacts, extending the capability for engineers to use concepts from the problem domain of discourse to specify apropos solutions. A key component in MDSD is domain-specific modeling languages (DSMLs) which are languages with focused expressiveness, targeting a specific taxonomy of problems. The de facto approach used is to first transform DSML models to an intermediate artifact in a HLL e.g., Java or C++, then execute that resulting code. Our research group has developed a class of DSMLs, referred to as interpreted DSMLs (i-DSMLs), where models are directly interpreted by a specialized execution engine with semantics based on model changes at runtime. This execution engine uses a layered architecture and is referred to as a domain-specific virtual machine (DSVM). As the domain-specific model being executed descends the layers of the DSVM the semantic gap between the user-defined model and the services being provided by the underlying infrastructure is closed. The focus of this research is the synthesis engine, the layer in the DSVM which transforms i-DSML models into executable scripts for the next lower layer to process. The appeal of an i-DSML is constrained as it possesses unique semantics contained within the DSVM. Existing DSVMs for i-DSMLs exhibit tight coupling between the implicit model of execution and the semantics of the domain, making it difficult to develop DSVMs for new i-DSMLs without a significant investment in resources. At the onset of this research only one i-DSML had been created for the user- centric communication domain using the aforementioned approach. This i-DSML is the Communication Modeling Language (CML) and its DSVM is the Communication Virtual machine (CVM). A major problem with the CVM’s synthesis engine is that the domain-specific knowledge (DSK) and the model of execution (MoE) are tightly interwoven consequently subsequent DSVMs would need to be developed from inception with no reuse of expertise. This dissertation investigates how to decouple the DSK from the MoE and sub- sequently producing a generic model of execution (GMoE) from the remaining appli- cation logic. This GMoE can be reused to instantiate synthesis engines for DSVMs in other domains. The generalized approach to developing the model synthesis com- ponent of i-DSML interpreters utilizes a reusable framework loosely coupled to DSK as swappable framework extensions. This approach involves first creating an i-DSML and its DSVM for a second do- main, demand-side smartgrid, or microgrid energy management, and designing the synthesis engine so that the DSK and MoE are easily decoupled. To validate the utility of the approach, the SEs are instantiated using the GMoE and DSKs of the two aforementioned domains and an empirical study to support our claim of reduced developmental effort is performed.

Transactions and schema evolution in a persistent object-oriented programming system

Applications are subject of a continuous evolution process with a profound impact on their underlining data model, hence requiring frequent updates in the applications' class structure and database structure as well. This twofold problem, schema evolution and instance adaptation, usually known as database evolution, is addressed in this thesis. Additionally, we address concurrency and error recovery problems with a novel meta-model and its aspect-oriented implementation. Modern object-oriented databases provide features that help programmers deal with object persistence, as well as all related problems such as database evolution, concurrency and error handling. In most systems there are transparent mechanisms to address these problems, nonetheless the database evolution problem still requires some human intervention, which consumes much of programmers' and database administrators' work effort. Earlier research works have demonstrated that aspect-oriented programming (AOP) techniques enable the development of flexible and pluggable systems. In these earlier works, the schema evolution and the instance adaptation problems were addressed as database management concerns. However, none of this research was focused on orthogonal persistent systems. We argue that AOP techniques are well suited to address these problems in orthogonal persistent systems. Regarding the concurrency and error recovery, earlier research showed that only syntactic obliviousness between the base program and aspects is possible. Our meta-model and framework follow an aspect-oriented approach focused on the object-oriented orthogonal persistent context. The proposed meta-model is characterized by its simplicity in order to achieve efficient and transparent database evolution mechanisms. Our meta-model supports multiple versions of a class structure by applying a class versioning strategy. Thus, enabling bidirectional application compatibility among versions of each class structure. That is to say, the database structure can be updated because earlier applications continue to work, as well as later applications that have only known the updated class structure. The specific characteristics of orthogonal persistent systems, as well as a metadata enrichment strategy within the application's source code, complete the inception of the meta-model and have motivated our research work. To test the feasibility of the approach, a prototype was developed. Our prototype is a framework that mediates the interaction between applications and the database, providing them with orthogonal persistence mechanisms. These mechanisms are introduced into applications as an {\it aspect} in the aspect-oriented sense. Objects do not require the extension of any super class, the implementation of an interface nor contain a particular annotation. Parametric type classes are also correctly handled by our framework. However, classes that belong to the programming environment must not be handled as versionable due to restrictions imposed by the Java Virtual Machine. Regarding concurrency support, the framework provides the applications with a multithreaded environment which supports database transactions and error recovery. The framework keeps applications oblivious to the database evolution problem, as well as persistence. Programmers can update the applications' class structure because the framework will produce a new version for it at the database metadata layer. Using our XML based pointcut/advice constructs, the framework's instance adaptation mechanism is extended, hence keeping the framework also oblivious to this problem. The potential developing gains provided by the prototype were benchmarked. In our case study, the results confirm that mechanisms' transparency has positive repercussions on the programmer's productivity, simplifying the entire evolution process at application and database levels. The meta-model itself also was benchmarked in terms of complexity and agility. Compared with other meta-models, it requires less meta-object modifications in each schema evolution step. Other types of tests were carried out in order to validate prototype and meta-model robustness. In order to perform these tests, we used an OO7 small size database due to its data model complexity. Since the developed prototype offers some features that were not observed in other known systems, performance benchmarks were not possible. However, the developed benchmark is now available to perform future performance comparisons with equivalent systems. In order to test our approach in a real world scenario, we developed a proof-of-concept application. This application was developed without any persistence mechanisms. Using our framework and minor changes applied to the application's source code, we added these mechanisms. Furthermore, we tested the application in a schema evolution scenario. This real world experience using our framework showed that applications remains oblivious to persistence and database evolution. In this case study, our framework proved to be a useful tool for programmers and database administrators. Performance issues and the single Java Virtual Machine concurrent model are the major limitations found in the framework.

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.

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.

vMotion Migración en Caliente

Esta memoria nos introduce en el mundo de los servidores de máquinas virtuales procurando no saltarse ningún paso, de una forma gráfica y sin necesidad de conocimientos previos sobre virtualización. Es una guía para la instalación y configuración de un centro de datos con las siguientes tecnologías de virtualización de la compañía VMware: vCenter Server y vSphere ESXi; y el almacenamiento en red open-source FreeNAS. Este despliegue se usará para poner a prueba el funcionamiento de la tecnología vMotion. vMotion es una tecnología para migrar en caliente una máquina virtual de un servidor de máquinas virtuales a otro, de forma transparente y sin desconexiones. Esta tecnología, con la potencia de los procesadores y el ancho de banda actual, es casi inocua al rendimiento de la máquina virtual, lo cual permite su aplicación en una gran diversidad de sectores.