Search

International audience; Today, the use of software is generally regulated by licenses, whether they are free or paid and with or without access to their sources. The world of licenses is very vast and unknown. Often only the public version is known (a software purchase corresponds to a license). For enterprises, the reality is much more complex, especially for main software publishers. Very few, if any, deployment algorithm takes Software Asset Management (SAM) considerations into account when placing software on Cloud architecture. This could have huge financial impact on the company using theses software. In this article, we present the SAM problem more deeply, then, after expressing our problem mathematically, we present GreenSAM, our multi-parametric heuristic handling performance and energy parameters as well as SAM considerations. We will then show the use of this heuristic on two realistic situations, first with an Oracle Database deployment and second with a larger scenario of managing a small OpenStack platform deployment. In both cases, we will compare GreenSAM with other heuristics to show how it handles the performance/energy criteria and the SAM compliance.

Microservices promise the benefits of services with an efficient granularity using dynamically allocated resources. In the current evolving architectures, data producers and consumers are created as decoupled components that support different data objects and quality of service. Actual implementations of service meshes lack support for data-driven paradigms, and focus on goal-based approaches designed to fulfill the general system goal. This diversity of available components demands the integration of users requirements and data products into the discovery mechanism. This paper proposes a data-driven service discovery framework based on profile matching using data-centric service descriptions. We have designed and evaluated a microservices architecture for providing service meshes with a standalone set of components that manages data profiles and resources allocations over multiple geographical zones. Moreover, we demonstrated an adaptation scheme to provide quality of service guarantees. Evaluation of the implementation on a real life testbed shows effectiveness of this approach with stable and fluctuating request incoming rates.

Over the past years, the Cloud phenomenon had an impressive increase in popularity in both the software industry and research worlds. The most interesting feature that Cloud Computing brings, from a Cloud client's point of view, is the on-demand resource provisioning model. This allows Cloud client platforms to be scaled up in order to accommodate more incoming clients and to scale down when the platform has unused resources, and this can all be done while the platform is running. As a result, the physical resources are used more efficiently and the Cloud client saves expenses. Achieving the above mentioned is not trivial and is done by leveraging more direct functionalities that Clouds provide. Three of these key functionalities are automatic scaling, load balancing and monitoring. They represent the focus of the current work. This report is divided into three main parts, one for each of the three main topics of interest: auto-scaling, load balancing and monitoring. We detail each of the three topics and present details on their presence and implementation in the commercial, open-source and research worlds. Among the commercial Cloud platform providers we have focused on Amazon EC2, Microsoft Azure, GoGrid and RackSpace. From the open-source initiatives we have focused our attention on Nimbus, Eucalyptus and OpenNebula. We have also detailed endeavours in the research world that are relevant to the topic of interest and are not connected to any of the Cloud providers. By studying this report, the reader will be able to understand the role that each of the three Cloud features plays for a Cloud client, will understand their inner workings and will be aware of the state-of-the-art available in current commercial and open-source Cloud platforms.

Cloud platforms have emerged as a leading solution for computation. In the meantime, large computations have shifted from big parallel tasks to workflows of smaller tasks with data dependencies between them. Task placement is a major issue on Cloud platforms, especially considering the impact of data exchanges on cost and makespan. In this paper, we investigate the consequences of network contention regarding the use of existing scheduling policies on DaaS-based platforms (DaaS for Data as a Service). We show here that the legacy algorithms use inefficient network models. We then modify those algorithms using a new model inspired by DaaS-based Cloud platforms. Thus, we manage to statically pack tasks so that a batch scheduler could deploy many real-time submitted workflows on a dynamic Cloud platform. Simulations of Fork-Join workflows deployment using SimGrid show that our algorithm reduces computation time as well as deployment costs.

An effective solution to deal with this dynamic nature of distributed systems is to implement a self-adaptive mechanism to sustain the distributed architecture. Self-adaptive systems can autonomously modify their behavior at run-timein response to changes in their environment. Our paper describes the self-adaptive algorithm that we developed for an existing middleware. Once the middleware is deployed, it can detects a set of events which indicate an unstable deployment state. When an event is detected, some instructions are executed to handle the event. We have proposed a sketch proof of the self-stabilizing property of the algorithm. We have designed a simulator to have a deeper insights of our proposed self-adaptive algorithm. Results of our simulated experiments validate the safe convergence of the algorithm. RÉSUMÉ.Dans cet article, nous nous intéressons aux moyens de rendre le déploiement d’un intergiciel auto-adaptatif. Le type d’intergiciel que nous avons considéré ici est hiérarchique (structure de graphe) et distribué. Les infrastructures de grilles/cloud étant dynamiques (perte et ajout de nœuds),un déploiement statique n’est pas la solution idéale car en cas de panne, il est souvent nécessaire de reprendre tout le processus de déploiement; or cette opération est très coûteuse. Nous avons donc proposé un algorithme auto-stabilisant pour que l’intergiciel puisse retrouver un état stable sans intervention extérieure, au bout d’un temps fini, lorsqu’il est confronté à des pannes transitoires. Pouravoir une idée plus précise des caractéristiques de l’algorithme, nous avons conçu un simulateur. Lesrésultats des simulations montrent qu’un déploiement, sujet à des pannes transitoires, s’auto-adapte.

Infrastructure as a Service clouds are a flexible and fast way to obtain (virtual) resources as demand varies. Grids, on the other hand, are middleware platforms able to combine resources from different administrative domains for task execution. Clouds can be used by grids as providers of devices such as virtual machines, so they only use the resources they need. But this requires grids to be able to decide when to allocate and release those resources. Here we introduce and analyze by simulations an economic mechanism (a) to set resource prices and (b) resolve when to scale resources depending on the users’ demand. This system has a strong emphasis on fairness, so no user hinders the execution of other users’ tasks by getting too many resources. Our simulator is based on the well-known GridSim software for grid simulation, which we expand to simulate infrastructure clouds. The results show how the proposed system can successfully adapt the amount of allocated resources to the demand, while at the same time ensuring that resources are fairly shared among users.

Mobile ad hoc networks as well as grid platforms are distributed, changing, and error prone environments. Communication costs within such infrastructure can be improved, or at least bounded, by using k-clustering. A k-clustering of a graph, is a partition of the nodes into disjoint sets, called clusters, in which every node is distance at most k from a designated node in its cluster, called the clusterhead. A self-stabilizing asynchronous distributed algorithm is given for constructing a k-clustering of a connected network of processes with unique IDs and weighted edges. The algorithm is comparison based, takes O(nk) time, and uses O(logn+logk) space per process, where n is the size of the network. To the best of our knowledge, this is the first solution to the k-clustering problem on weighted graphs.

Best Paper Award; International audience; The adoption of cloud computing is still limited by several legal concerns from companies. One of those reasons is the data sovereignty, as data can be physically host in sensible locations, resulting in a lack of control for companies. In this paper, we present the Nu@ge project aimed at building a federation of container-sized datacenter on the French territory. Nu@ge provides a software stack that enables companies to put independent datacenters in cooperation in a national mesh. Additionally, a prototype of a container-sized datacenter has been validated and patented.

International audience; Within computational Grids, some services (typically software components, e.g., linear algebra libraries) are made available by some servers to some clients. In spite of the growing popularity of such Grids, the service discovery, although efficient in many cases, does not reach several requirements. Among them, the flexibility of the discovery and its efficiency on wide-area dynamic platforms are two major issues. Therefore, it becomes crucial to propose new tools coping with such platforms. Emerging peer-to-peer technologies provide algorithms allowing the distribution and the retrieval of data items while addressing the dynamicity of the underlying network. Whereas merging peer-to-peer technology and Grid infrastructures has been widely suggested, very few implementations are available. The contribution of this paper is twofold. First, we present the design, the implementation and the experimentation of the first architecture, to our knowledge, extending traditional Network-Enabled Servers (NES) systems with an unstructured peer-to-peer network. This extension allows to dynamically connect distributed agents thus providing to clients an entry point to servers geographically distributed. Our implementation is based on the Diet middleware and the JXTA toolbox and experimentation have been conducted on a high speed network. Then, we study the service discovery in a pure peer-to-peer environment. We describe a new trie-based approach for the peer-to-peer service discovery service, supporting range queries while providing fault-tolerance and taking into account the topology of the underlying network. We validate this approach both by analysis and simulation.

The adoption of cloud computing is still limited by several legal concerns that customers may have, such as data sovereignty. In cloud computing, data can be physically hosted in sensible locations , resulting in a lack of control for companies. In this context, we present the Nu@ge project, which aims at building a federation of container-sized datacenters in the French territory. Nu@ge provides a software stack that enables companies to interconnect independent datacenters in a national mesh. A software architecture is presented and implemented as a federation of small datacenters deployed in France. The proposed architecture enables cooperation between local customized-cloud managers and a federation-wide middleware. It uses monitoring information from facilities and performance indicators from physical servers for managing the system, preventing incidents and considering energy efficiency. Additionally, a prototype of a container-sized datacenter has been validated and patented. KEYWORDS cloud federation, energy-aware scheduling, IaaS 1 INTRODUCTION Cloud computing represents a significant evolution of information and communications technology in usage and organization. It allows companies to increase their competitiveness by lowering IT infrastructure costs and improving quality of service. This field is an opportuinity to new markets and business models as it is expected to grow up to 29% per year until 2019. 1 Despite its benefits to users, cloud computing raises several concerns of applications, data storage, and processing. Cloud providers reveal few information about geographical location and the way to process of data and applications. As information converted and stored in binary digital form is subject to the laws of the country in which it is located, several concerns are raised from a legal standpoint. Third-party entities or governments could take control of sensible data, and legal protections may not apply if one's data are located outside her country. Additionally, data from a company could be physically hosted with data from others. This causes security risks as one company may attempt to access data of another. The Nu@ge consortium* comprises 6 SMEs (small and medium-sized enterprise) and 2 research teams aiming to create cloud technology that is open to heterogeneous hardware and software stacks, spread on *Nu@ge is a research project funded by the FSN (Fund for the Digital Society, BPI France) as part of the Investissement d' Avenir program. http://www.nuage-france.fr a regional network among France and that uses low-energy consumption and ecological datacenters. This Nu@ge project is a large effort that tackles several challenges. In this work, we describe a federated architecture to provision virtual clusters of resources over the network while providing administrators with a control over data location and quality of service. Our solution is based on innovative container-sized datacenters that enables deployment of a cost-effective and high-performance datacenter environment in any location, meshing regional company-owned datacenters. This paper is structured as follows. Section 2 introduces the related work. Section 3 presents the Nu@ge architecture and the associated technical choices. The architecture is presented in Section 4 while the prototype is described in Section 5. Section 6 discusses an evaluation of energy efficiency, and Section 7 presents the energy-aware adaptive placement of tasks on server. Section 8 concludes the paper. 2 BACKGROUND AND RELATED WORK This section presents related work in the context of cloud providers and existing cloud federation approaches. Then, we detail the notion of modular datacenters and its expected features. Finally, we justify the choice of implementing our own distributed storage solution and the different energy-aware scheduling mechanisms.

International audience; Complying with security and privacy requirements of appliances such as mobile handsets, personal computers, servers for customers, enterprises and governments is mandatory to prevent from theft of sensitive data and to preserve their integrity. Nowadays, with the rising of the Cloud Computing approach in business fields, security and privacy are even more critical. The aim of this article is then to propose a way to build a secure and trustable Cloud. The idea is to spread and embed Secure Elements (SE) on each level of the Cloud in order to make a wide trusted infrastructure which complies with access control and isolation policies. This article presents therefore this new approach of trusted Cloud infrastructure based on a Network of Secure Elements (NoSE), and it illustrates this approach through different use cases.

Ability to find and get services is a key requirement in the development of large-scale distributed sys- tems. We consider dynamic and unstable environments, namely Peer-to-Peer (P2P) systems. In previous work, we designed a service discovery solution called Distributed Lexicographic Placement Table (DLPT), based on a hierar- chical overlay structure. A self-stabilizing version was given using the Propagation of Information with Feedback (PIF) paradigm. In this paper, we introduce the self-stabilizing COPIF (for Collaborative PIF) scheme. An algo- rithm is provided with its correctness proof. We use this approach to improve a distributed P2P framework designed for the services discovery. Significantly efficient experimental results are presented., Comment: (2012)

Many scientific applications are described through workflow structures. Due to the increasing level of parallelism offered by modern computing infrastructures, workflow applications now have to be composed not only of sequential programs, but also of parallel ones. Cloud platforms bring on-demand resource provisioning and pay-as-you-go billing model. Then the execution of a workflow corresponds to a certain budget. The current work addresses the problem of resource allocation for non-deterministic workflows under budget constraints. We present a way of transforming the initial problem into sub-problems that have been studied before. We propose two new allocation algorithms that are capable of determining resource allocations under budget constraints and we present ways of using them to address the problem at hand.

At times of large scale, heterogeneous computing infrastructures gathering myriads of services, service discovery has become a critical feature that has to deal with the scale and dynamic nature of such platforms. The SPADES project (Servicing Petascale Architectures and Distributed System) is a consortium whose purpose is to promote new solutions to deal with distributed, volatile and heterogeneous computing resources. For the aforementioned reasons, at the project's core, the service discovery has been envisioned as fully decentralized. More precisely, the proposed P2P service discovery system proposed within the SPADES project is based on the DLPT approach (Distributed Lexicographic Placement Table) providing distributed structures and algorithms for such a feature. In this paper, an implementation of the DLPT concepts into the middleware developed within the SPADES project, called SBAM, is devised. Moreover, its actual deployment over a nation-wide grid system, as well as its performance are detailed.

International audience; Accessing the power of distributed resources can nowadays easily be done using a middleware based on a client/server approach. Several architectures exist for those middleware's. The most scalable ones rely on a hierarchical design. Determining the best shape for the hierarchy, the one giving the best throughput of services, is not an easy task. We first propose a computation and communication model for such hierarchical middleware. Our model takes into account the deployment of several services in the hierarchy. Then, based on this model, we propose algorithms for automatically constructing a hierarchy on two kinds of heterogeneous platforms: communication homogeneous/computation heterogeneous platforms, and fully heterogeneous platforms. The proposed algorithms aim at offering the users the best obtained to requested throughput ratio, while providing fairness on this ratio for the different kinds of services, and using as few resources as possible for the hierarchy. For each kind of platforms, we compare our model with experimental results on a real middleware called DIET (Distributed Interactive Engineering Toolbox).

International audience; Within distributed computing platforms, some computing abilities (or services) are offered to clients. To build dynamic applications using such services as basic blocks, a critical prerequisite is to discover those services. Traditional approaches to the service discovery problem have historically relied upon centralized solutions, unable to scale well in large unreliable platforms. In this chapter, we will first give an overview of the state of the art of service discovery solutions based on peer-to-peer (P2P) technologies that allow such a functionality to remain efficient at large scale. We then focus on one of these approaches: the Distributed Lexicographic Placement Table (DLPT) architecture, that provide particular mechanisms for load balancing and fault-tolerance. This solution centers around three key points. First, it calls upon an indexing system structured as a prefix tree, allowing multi-attribute range queries. Second, it allows the mapping of such structures onto heterogeneous and dynamic networks and proposes some load balancing heuristics for it. Third, as our target platform is dynamic and unreliable, we describe its powerful fault-tolerance mechanisms, based on self-stabilization. Finally, we present the software prototype of this architecture and its early experiments.

International audience; Accessing the power of distributed resources can nowadays easily be done using a middleware based on a client/server approach. Several architectures exist for those middleware. The most scalable ones rely on a hierarchical design. Determining the best shape for the hierarchy, the one giving the best throughput of services, is not an easy task. We first propose a computation and communication model for such hierarchical middleware. Our model takes into account the deployment of several services in the hierarchy. Then, based on this model, we propose an algorithm for automatically constructing a hierarchy. This algorithm aims at offering the users the best obtained to requested throughput ratio, while providing fairness on this ratio for the different kind of services, and using as few resources as possible. Finally, we compare our model with experimental results on a real middleware called Diet.

International audience; SystemC has become a very popular standardized language for the modeling of system-on-chip (SoC) devices. However, due to the ever increasing complexity of SoC designs, the ever longer simulation times affect SoC exploration potential and time-to-market. In order to reduce these times, we have developed a parallel SystemC kernel. Because the SystemC semantics require a high level of synchronization which can dramatically affect the performance gains, we investigate in this paper some ways to reduce the synchronization overheads. We validate then our approaches against an academic design model and a real, industrial application.

The efficiency of service discovery is critical in the development of fully decentralized middleware intended to manage large scale computational grids. This demand influenced the design of many peer-to-peer based approaches. The ability to cope with the expressiveness of the service discovery was behind the design of a new kind of overlay structures that is based on tries, or prefix trees. Although these overlays are well designed, one of their weaknesses is the lack of any concrete fault tolerant mechanism, especially in dynamic platforms; the faults are handled by using preventive and costly mechanisms, \eg using a high degree of replication. Moreover, those systems cannot handle any arbitrary transient failure. Self-stabilization, which is an efficient approach to designreliable solutions for dynamic systems, was recently suggested to be a good alternative to inject fault-tolerance in peer-to-peer systems. However, most of the previous research on self-stabilization in tree and/or P2P networks was designed in theoretical models, making these approaches hard to implement in practice. In this paper, we provide a self-stabilizing message passing protocol to maintain prefix trees over practical peer-to-peer networks. A complete correctness proof is provided, as well as simulation results to estimate the practical impact of our protocol.

The efficiency of service discovery is a crucial point in the development of fully decentralized middlewares intended to manage large scale computational grids. The work conducted on this issue led to the design of many peer-to-peer fashioned approaches. More specifically, the need for flexibility and complexity in the service discovery has seen the emergence of a new kind of overlays, based on tries, also known as lexicographic trees. Although these overlays are efficient and well designed, they require a costly maintenance and do not accurately take into account the heterogeneity of nodes and the changing popularity of the services requested by users. In this paper, we focus on reducing the cost of the maintenance of a particular architecture, based on a dynamic prefix tree, while enhancing it with some load balancing techniques that dynamically adapt the load of the nodes in order to maximize the throughput of the system. The algorithms developed couple a self-organizing prefix tree overlay with load balancing techniques inspired by similar previous works undertaken for distributed hash tables. After some simulation results showing how our load balancing heuristics perform in such an overlay and compare to other heuristics, we provide a fair comparison of this architecture and similar overlays recently proposed.; L’efficacité de la découverte de services est un point crucial du développement d’intergiciels de grille totalement décentralisés. Les travaux ayant pour but la résolution de ce problème ont généré un certain nombre d’approches pair-à-pair. le besoin de flexibilité et d’expressivité a donné lieu au développement d’architecture s’appuyant sur des arbres de préfixes(ou arbres lexicographiques). Ces overlays souffrent d’une maintenance couteuse et ne prennent pas en compte la nature hétérogène de la plate-forme physique sous-jacente et la popularité différente et changeante de chaque ressource enregistrée.Dans ce rapport, nous nous focalisons sur la réduction du cout de maintenance d’une telle architecture, basée sur un arbre de préfixes dynamique,tout en lui donnant la possibilité de s’adapter à l’hétérogénéité précitée par l’enrichissant de mécanismes de répartition de la charge qui adaptent dynamiquement la charge des nœuds dans le but de maximiser le débit sur service. Notre approche couple des travaux de répartition de la charge dans les DHTs avec un overlay en arbre de préfixes auto-organisant. Après des résultats de simulation mettant en évidence l’efficacité de notre heuristique, nous comparons notre approche avec les travaux s’appuyant sur des structures distribuées similaires.

In this paper we present a model and simulator for many clusters of heterogeneous PCs belonging to a local network. These clusters are assumed to be connected to each other through a global network and each cluster is managed via a local scheduler which is shared by many users. We validate our simulator by comparing the experimental and analytical results of a M/M/4 queuing system. These studies indicate that the simulator is consistent. Next, we do the comparison with a real batch system and we obtain an average error of 10.5% for the response time and 12% for the makespan. We conclude that the simulator is realistic and well as the behaviour of a large-scale system. Thus we can study the scheduling of our system called DIRAC in a high throughput context. We justify our decentralized, adaptive and opportunistic approach in comparison to a centralized approach in such a context

Large scale distributed systems like Grids are difficult to study only from theoretical models and simulators. Most Grids deployed at large scale are production platforms that are inappropriate research tools because of their limited reconfiguration, control and monitoring capabilities. In this paper, we present Grid'5000, a 5000 CPUs nation-wide infrastructure for research in Grid computing. Grid'5000 is designed to provide a scientific tool for computer scientists similar to the large-scale instruments used by physicists, astronomers and biologists. We describe the motivations, design, architecture, configuration examples of Grid'5000 and performance results for the reconfiguration subsystem.

Summary form only given. This article focuses on the deployment of grid infrastructures, more specifically problem solving environments (PSE) for numerical applications on the grid. Although the deployment of such an architecture may be constrained e.g., firewall, right access or security, its efficiency heavily depends on the quality of the mapping between its different components and the grid resources. This article proposes a new model based on linear programming to estimate the performance of a deployment of a hierarchical PSE. The advantages of our modeling approach are: evaluate a virtual deployment before a real deployment, provide a decision builder tool (i.e., designed to compare different architectures or add new resources) and take into account the platform scalability. Using our model, it is possible to determine the bottleneck of the platform and thus to know whether a given deployment can be improved or not. We illustrate the model by applying the results to improve performance of an existing hierarchical PSE called DIET.

In this paper, we present an analytical performance model of the parallel left-right looking out-of-core LU factorization algorithm. We show the accuracy of the performance prediction for a prototype implementation in the ScaLAPACK library. We will show that with a correct distribution of the matrix and with an overlapof IO by computation, we obtain performances similar to those of the in-core algorithm. To get such performances, the size of the physical main memory only need to be proportional to the product of the matrix order (not the matrix size) by the ratio of the IO bandwidth and the computation rate: There is no need of large main memory for the factorization of huge matrix!

This paper presents a basic implementation of the OGF standard GridRPC Data Management API and its integration in two different middleware, respectively DIET and Ninf. We have conducted several experiments, showing the benefits a Grid user can expect 1) in terms of computation feasibility and resource usage compared to the current GridRPC context since useless transfers can be avoided; 2) in terms of reducing the completion time of an application to obtain results the soonest (data and temporary results can be easily kept on the computational and storage servers); 3) in terms of code portability, since we show with these examples that at last the same GridRPC code can be compiled and executed within two different GridRPC middleware which implements the GridRPC data management API; 4) finally we thus obtain middleware interoperability without any explicit glue: we show as a proof of concept that resources dispatched across different administrative domains can be used altogether without the underlying distributed data management but this needs knowledge of the topology and/or computing resources: computational servers of DIET and Ninf transparently collaborate to the same calculus by sharing GridRPC data!

International audience; As grids become more and more attractive for solving complex problems with high computational and storage requirements, the need for adequate grid programming models is considerable. To this purpose, the GridRP C model has been proposed as a grid version of the classical RPC paradigm, with the goal to build NES (Network-Enabled Server) environments. In this model, data management has not been defined and is now explicitly left at the user's charge. The contribution of this paper is to enhance data management in NES by introducing a transparent data access model, available through the concept of grid data-sharing service. Data management is totally delegated to the service, whereas the applications simply access shared data via global identifiers. We illustrate our approach using the DIET GridRP C middleware and the J UXME M data-sharing service. Notably, our experiments performed on the Grid'5000 using a real-life application show the efficiency of using JuxMem for managing persistent data in the GridRPC model: application execution times in a grid environment are of the same order as in a cluster environment.