Tolerância a falha em um ambiente de computação em nuvem open source

Pós-graduação em Ciência da Computação - IBILCE

Strumenti di gestione per infrastrutture IT

L'elaborato analizza gli strumenti utilizzati nella gestione di infrastrutture informatiche. In particolare: nel primo capitolo viene fatta una panoramica del Cloud Computing e relativi modelli infrastrutturali e di distribuzione. Nel secondo vengono analizzate le piattaforme di gestione di sistemi basati su architettura IaaS, quali OpenStack, OpenNebula ed Eucalyptus. Nel terzo capitolo vengono esaminati due dei Software Configuration Management maggiormente utilizzati: Puppet e Chef. Nel quarto ed ultimo capitolo viene descritto lo standard OCCI (Open Cloud Computing Interface).

Composizione dinamica di funzioni di rete virtuali in ambienti cloud

Questo documento si interroga sulle nuove possibilità offerte agli operatori del mondo delle Reti di Telecomunicazioni dai paradigmi di Network Functions Virtualization, Cloud Computing e Software Defined Networking: questi sono nuovi approcci che permettono la creazione di reti dinamiche e altamente programmabili, senza disdegnare troppo il lato prestazionale. L'intento finale è valutare se con un approccio di questo genere si possano implementare dinamicamente delle concatenazioni di servizi di rete e se le prestazioni finali rispecchiano ciò che viene teorizzato dai suddetti paradigmi. Tutto ciò viene valutato per cercare una soluzione efficace al problema dell'ossificazione di Internet: infatti le applicazioni di rete, dette middle-boxes, comportano costi elevati, situazioni di dipendenza dal vendor e staticità delle reti stesse, portando all'impossibilità per i providers di sviluppare nuovi servizi. Il caso di studio si basa proprio su una rete che implementa questi nuovi paradigmi: si farà infatti riferimento a due diverse topologie, una relativa al Livello L2 del modello OSI (cioè lo strato di collegamento) e una al Livello L3 (strato di rete). Le misure effettuate infine mostrano come le potenzialità teorizzate siano decisamente interessanti e innovative, aprendo un ventaglio di infinite possibilità per il futuro sviluppo di questo settore.

Migrazione concorrente di macchine virtuali su piattaforme open source

Questa tesi si pone l’obiettivo di effettuare un’analisi aggiornata sulla recente evoluzione del Cloud Computing e dei nuovi modelli architetturali a sostegno della continua crescita di richiesta di risorse di computazione, di storage e di rete all'interno dei data center, per poi dedicarsi ad una fase sperimentale di migrazioni live singole e concorrenti di macchine virtuali, studiandone le prestazioni a livello di risorse applicative e di rete all’interno della piattaforma open source di virtualizzazione QEMU-KVM, oggi alla base di sistemi cloud-based come Openstack. Nel primo capitolo, viene effettuato uno studio dello stato dell’arte del Cloud Computing, dei suoi attuali limiti e delle prospettive offerte da un modello di Cloud Federation nel futuro immediato. Nel secondo capitolo vengono discusse nel dettaglio le tecniche di live migration, di recente riferimento per la comunità scientifica internazionale e le possibili ottimizzazioni in scenari inter e intra data center, con l’intento di definire la base teorica per lo studio approfondito dell’implementazione effettiva del processo di migrazione su piattaforma QEMU-KVM, che viene affrontato nel terzo capitolo. In particolare, in quest’ultimo sono descritti i principi architetturali e di funzionamento dell'hypervisor e viene definito il modello di progettazione e l’algoritmo alla base del processo di migrazione. Nel quarto capitolo, infine, si presenta il lavoro svolto, le scelte configurative e progettuali per la creazione di un ambiente di testbed adatto allo studio di sessioni di live migration concorrenti e vengono discussi i risultati delle misure di performance e del comportamento del sistema, tramite le sperimentazioni effettuate.

Virtualizzazione di funzioni di rete su piattaforme per cloud computing

Questo documento affronta le novità ed i vantaggi introdotti nel mondo delle reti di telecomunicazioni dai paradigmi di Software Defined Networking e Network Functions Virtualization, affrontandone prima gli aspetti teorici, per poi applicarne i concetti nella pratica, tramite casi di studio gradualmente più complessi. Tali innovazioni rappresentano un'evoluzione dell'architettura delle reti predisposte alla presenza di più utenti connessi alle risorse da esse offerte, trovando quindi applicazione soprattutto nell'emergente ambiente di Cloud Computing e realizzando in questo modo reti altamente dinamiche e programmabili, tramite la virtualizzazione dei servizi di rete richiesti per l'ottimizzazione dell'utilizzo di risorse. Motivo di tale lavoro è la ricerca di soluzioni ai problemi di staticità e dipendenza, dai fornitori dei nodi intermedi, della rete Internet, i maggiori ostacoli per lo sviluppo delle architetture Cloud. L'obiettivo principale dello studio presentato in questo documento è quello di valutare l'effettiva convenienza dell'applicazione di tali paradigmi nella creazione di reti, controllando in questo modo che le promesse di aumento di autonomia e dinamismo vengano rispettate. Tale scopo viene perseguito attraverso l'implementazione di entrambi i paradigmi SDN e NFV nelle sperimentazioni effettuate sulle reti di livello L2 ed L3 del modello OSI. Il risultato ottenuto da tali casi di studio è infine un'interessante conferma dei vantaggi presentati durante lo studio teorico delle innovazioni in analisi, rendendo esse una possibile soluzione futura alle problematiche attuali delle reti.

Tecniche Cloud Avanzate per Adaptive Hadoop Load Balancing tramite Sahara

Ogni giorno vengono generati grandi moli di dati attraverso sorgenti diverse. Questi dati, chiamati Big Data, sono attualmente oggetto di forte interesse nel settore IT (Information Technology). I processi digitalizzati, le interazioni sui social media, i sensori ed i sistemi mobili, che utilizziamo quotidianamente, sono solo un piccolo sottoinsieme di tutte le fonti che contribuiscono alla produzione di questi dati. Per poter analizzare ed estrarre informazioni da questi grandi volumi di dati, tante sono le tecnologie che sono state sviluppate. Molte di queste sfruttano approcci distribuiti e paralleli. Una delle tecnologie che ha avuto maggior successo nel processamento dei Big Data, e Apache Hadoop. Il Cloud Computing, in particolare le soluzioni che seguono il modello IaaS (Infrastructure as a Service), forniscono un valido strumento all'approvvigionamento di risorse in maniera semplice e veloce. Per questo motivo, in questa proposta, viene utilizzato OpenStack come piattaforma IaaS. Grazie all'integrazione delle tecnologie OpenStack e Hadoop, attraverso Sahara, si riesce a sfruttare le potenzialita offerte da un ambiente cloud per migliorare le prestazioni dell'elaborazione distribuita e parallela. Lo scopo di questo lavoro e ottenere una miglior distribuzione delle risorse utilizzate nel sistema cloud con obiettivi di load balancing. Per raggiungere questi obiettivi, si sono rese necessarie modifiche sia al framework Hadoop che al progetto Sahara.

A Case for CDN-as-a-Service in the Cloud: A Mobile Cloud Networking Argument

Content Distribution Networks are mandatory components of modern web architectures, with plenty of vendors offering their services. Despite its maturity, new paradigms and architecture models are still being developed in this area. Cloud Computing, on the other hand, is a more recent concept which has expanded extremely quickly, with new services being regularly added to cloud management software suites such as OpenStack. The main contribution of this paper is the architecture and the development of an open source CDN that can be provisioned in an on-demand, pay-as-you-go model thereby enabling the CDN as a Service paradigm. We describe our experience with integration of CDNaaS framework in a cloud environment, as a service for enterprise users. We emphasize the flexibility and elasticity of such a model, with each CDN instance being delivered on-demand and associated to personalized caching policies as well as an optimized choice of Points of Presence based on exact requirements of an enterprise customer. Our development is based on the framework developed in the Mobile Cloud Networking EU FP7 project, which offers its enterprise users a common framework to instantiate and control services. CDNaaS is one of the core support components in this project as is tasked to deliver different type of multimedia content to several thousands of users geographically distributed. It integrates seamlessly in the MCN service life-cycle and as such enjoys all benefits of a common design environment, allowing for an improved interoperability with the rest of the services within the MCN ecosystem.

Demo -- Closer to Cloud-RAN: RAN as a Service

Commoditization and virtualization of wireless networks are changing the economics of mobile networks to help network providers (e.g., MNO, MVNO) move from proprietary and bespoke hardware and software platforms toward an open, cost-effective, and flexible cellular ecosystem. In addition, rich and innovative local services can be efficiently created through cloudification by leveraging the existing infrastructure. In this work, we present RANaaS, which is a cloudified radio access network delivered as a service. RANaaS provides the service life-cycle of an ondemand, elastic, and pay as you go 3GPP RAN instantiated on top of the cloud infrastructure. We demonstrate an example of realtime cloudified LTE network deployment using the OpenAirInterface LTE implementation and OpenStack running on commodity hardware as well as the flexibility and performance of the platform developed.

Mobility and Bandwidth Prediction as a Service in Virtualized LTE Systems

Recently telecommunication industry benefits from infrastructure sharing, one of the most fundamental enablers of cloud computing, leading to emergence of the Mobile Virtual Network Operator (MVNO) concept. The most momentous intents by this approach are the support of on-demand provisioning and elasticity of virtualized mobile network components, based on data traffic load. To realize it, during operation and management procedures, the virtualized services need be triggered in order to scale-up/down or scale-out/in an instance. In this paper we propose an architecture called MOBaaS (Mobility and Bandwidth Availability Prediction as a Service), comprising two algorithms in order to predict user(s) mobility and network link bandwidth availability, that can be implemented in cloud based mobile network structure and can be used as a support service by any other virtualized mobile network services. MOBaaS can provide prediction information in order to generate required triggers for on-demand deploying, provisioning, disposing of virtualized network components. This information can be used for self-adaptation procedures and optimal network function configuration during run-time operation, as well. Through the preliminary experiments with the prototype implementation on the OpenStack platform, we evaluated and confirmed the feasibility and the effectiveness of the prediction algorithms and the proposed architecture.

A Demonstration of Mobility Prediction as a Service in Cloudified LTE Networks

Location prediction has attracted a significant amount of research effort. Being able to predict users’ movement benefits a wide range of communication systems, including location-based service/applications, mobile access control, mobile QoS provision, and resource management for mobile computation and storage management. In this demo, we present MOBaaS, which is a cloudified Mobility and Bandwidth prediction services that can be instantiated, deployed, and disposed on-demand. Mobility prediction of MOBaaS provides location predictions of a single/group user equipments (UEs) in a future moment. This information can be used for self-adaptation procedures and optimal network function configuration during run-time operations. We demonstrate an example of real-time mobility prediction service deployment running on OpenStack platform, and the potential benefits it bring to other invoking services.

An architecture for a heterogeneous private IaaS management system

Cloud computing and, more particularly, private IaaS, is seen as a mature technology with a myriad solutions tochoose from. However, this disparity of solutions and products has instilled in potential adopters the fear of vendor and data lock-in. Several competing and incompatible interfaces and management styles have given even more voice to these fears. On top of this, cloud users might want to work with several solutions at the same time, an integration that is difficult to achieve in practice. In this paper, we propose a management architecture that tries to tackle these problems; it offers a common way of managing several cloud solutions, and an interface that can be tailored to the needs of the user. This management architecture is designed in a modular way, and using a generic information model. We have validated our approach through the implementation of the components needed for this architecture to support a sample private IaaS solution: OpenStack

Development of a management system for virtual machines on private clouds

Cloud computing and, more particularly, private IaaS, is seen as a mature technol- ogy with a myriad solutions to choose from. However, this disparity of solutions and products has instilled in potential adopters the fear of vendor and data lock- in. Several competing and incompatible interfaces and management styles have increased even more these fears. On top of this, cloud users might want to work with several solutions at the same time, an integration that is difficult to achieve in practice. In this Master Thesis I propose a management architecture that tries to solve these problems; it provides a generalized control mechanism for several cloud infrastructures, and an interface that can meet the requirements of the users. This management architecture is designed in a modular way, and using a generic infor- mation model. I have validated the approach through the implementation of the components needed for this architecture to support a sample private IaaS solution: OpenStack.

Interconexión de centros de datos mediante técnicas SDN

En este trabajo final de grado se ha contribuido a la interconexión de centros de datos distribuidos geográficamente, integrando para ello nuevas funcionalidades en la arquitectura Application-Based Network Operations (ABNO) y configurando los componentes software necesarios. ABNO engloba distintas tecnologías que recogen la información sobre los recursos disponibles en la red con el objetivo de proporcionar rutas específicas para el tráfico. La solución que se presenta en este trabajo se basa en las redes definidas por software (Software-Defined Networking, SDN), como solución innovadora para mejorar la gestión y el control de las infraestructuras que pertenecen a múltiples dominios administrativos, pero trabajan en colaboración en una federación común mejorando la calidad del servicio ofrecido. La conectividad entre los diferentes dominios es posible gracias a los Túneles GRE. Cada centro de datos supone un dominio administrativo diferenciado, disponiendo cada uno de ellos del software de gestión en la nube OpenStack para la creación de las máquinas virtuales (VM) que posteriormente serán interconectadas. Además, cada centro de datos también contará con el controlador Ryu SDN que se encargará del control de la conectividad, siendo también independiente para cada uno de estos dominios. Con el objetivo de mantener una visión integral de todos los recursos de la red disponibles, y de proporcionar una conectividad extremo a extremo (E2E) requerida por los centros de datos, la arquitectura ABNO ha tenido que ser modificada para soportar estas nuevas funcionalidades, así como validada en un escenario con infraestructuras multidominio.

A Novel Approach to Context-Aware Content-Based Middleware

La computación ubicua está extendiendo su aplicación desde entornos específicos hacia el uso cotidiano; el Internet de las cosas (IoT, en inglés) es el ejemplo más brillante de su aplicación y de la complejidad intrínseca que tiene, en comparación con el clásico desarrollo de aplicaciones. La principal característica que diferencia la computación ubicua de los otros tipos está en como se emplea la información de contexto. Las aplicaciones clásicas no usan en absoluto la información de contexto o usan sólo una pequeña parte de ella, integrándola de una forma ad hoc con una implementación específica para la aplicación. La motivación de este tratamiento particular se tiene que buscar en la dificultad de compartir el contexto con otras aplicaciones. En realidad lo que es información de contexto depende del tipo de aplicación: por poner un ejemplo, para un editor de imágenes, la imagen es la información y sus metadatos, tales como la hora de grabación o los ajustes de la cámara, son el contexto, mientras que para el sistema de ficheros la imagen junto con los ajustes de cámara son la información, y el contexto es representado por los metadatos externos al fichero como la fecha de modificación o la de último acceso. Esto significa que es difícil compartir la información de contexto, y la presencia de un middleware de comunicación que soporte el contexto de forma explícita simplifica el desarrollo de aplicaciones para computación ubicua. Al mismo tiempo el uso del contexto no tiene que ser obligatorio, porque si no se perdería la compatibilidad con las aplicaciones que no lo usan, convirtiendo así dicho middleware en un middleware de contexto. SilboPS, que es nuestra implementación de un sistema publicador/subscriptor basado en contenido e inspirado en SIENA [11, 9], resuelve dicho problema extendiendo el paradigma con dos elementos: el Contexto y la Función de Contexto. El contexto representa la información contextual propiamente dicha del mensaje por enviar o aquella requerida por el subscriptor para recibir notificaciones, mientras la función de contexto se evalúa usando el contexto del publicador y del subscriptor. Esto permite desacoplar la lógica de gestión del contexto de aquella de la función de contexto, incrementando de esta forma la flexibilidad de la comunicación entre varias aplicaciones. De hecho, al utilizar por defecto un contexto vacío, las aplicaciones clásicas y las que manejan el contexto pueden usar el mismo SilboPS, resolviendo de esta forma la incompatibilidad entre las dos categorías. En cualquier caso la posible incompatibilidad semántica sigue existiendo ya que depende de la interpretación que cada aplicación hace de los datos y no puede ser solucionada por una tercera parte agnóstica. El entorno IoT conlleva retos no sólo de contexto, sino también de escalabilidad. La cantidad de sensores, el volumen de datos que producen y la cantidad de aplicaciones que podrían estar interesadas en manipular esos datos está en continuo aumento. Hoy en día la respuesta a esa necesidad es la computación en la nube, pero requiere que las aplicaciones sean no sólo capaces de escalar, sino de hacerlo de forma elástica [22]. Desgraciadamente no hay ninguna primitiva de sistema distribuido de slicing que soporte un particionamiento del estado interno [33] junto con un cambio en caliente, además de que los sistemas cloud actuales como OpenStack u OpenNebula no ofrecen directamente una monitorización elástica. Esto implica que hay un problema bilateral: cómo puede una aplicación escalar de forma elástica y cómo monitorizar esa aplicación para saber cuándo escalarla horizontalmente. E-SilboPS es la versión elástica de SilboPS y se adapta perfectamente como solución para el problema de monitorización, gracias al paradigma publicador/subscriptor basado en contenido y, a diferencia de otras soluciones [5], permite escalar eficientemente, para cumplir con la carga de trabajo sin sobre-provisionar o sub-provisionar recursos. Además está basado en un algoritmo recientemente diseñado que muestra como añadir elasticidad a una aplicación con distintas restricciones sobre el estado: sin estado, estado aislado con coordinación externa y estado compartido con coordinación general. Su evaluación enseña como se pueden conseguir notables speedups, siendo el nivel de red el principal factor limitante: de hecho la eficiencia calculada (ver Figura 5.8) demuestra cómo se comporta cada configuración en comparación con las adyacentes. Esto permite conocer la tendencia actual de todo el sistema, para saber si la siguiente configuración compensará el coste que tiene con la ganancia que lleva en el throughput de notificaciones. Se tiene que prestar especial atención en la evaluación de los despliegues con igual coste, para ver cuál es la mejor solución en relación a una carga de trabajo dada. Como último análisis se ha estimado el overhead introducido por las distintas configuraciones a fin de identificar el principal factor limitante del throughput. Esto ayuda a determinar la parte secuencial y el overhead de base [26] en un despliegue óptimo en comparación con uno subóptimo. Efectivamente, según el tipo de carga de trabajo, la estimación puede ser tan baja como el 10 % para un óptimo local o tan alta como el 60 %: esto ocurre cuando se despliega una configuración sobredimensionada para la carga de trabajo. Esta estimación de la métrica de Karp-Flatt es importante para el sistema de gestión porque le permite conocer en que dirección (ampliar o reducir) es necesario cambiar el despliegue para mejorar sus prestaciones, en lugar que usar simplemente una política de ampliación. ABSTRACT The application of pervasive computing is extending from field-specific to everyday use. The Internet of Things (IoT) is the shiniest example of its application and of its intrinsic complexity compared with classical application development. The main characteristic that differentiates pervasive from other forms of computing lies in the use of contextual information. Some classical applications do not use any contextual information whatsoever. Others, on the other hand, use only part of the contextual information, which is integrated in an ad hoc fashion using an application-specific implementation. This information is handled in a one-off manner because of the difficulty of sharing context across applications. As a matter of fact, the application type determines what the contextual information is. For instance, for an imaging editor, the image is the information and its meta-data, like the time of the shot or camera settings, are the context, whereas, for a file-system application, the image, including its camera settings, is the information and the meta-data external to the file, like the modification date or the last accessed timestamps, constitute the context. This means that contextual information is hard to share. A communication middleware that supports context decidedly eases application development in pervasive computing. However, the use of context should not be mandatory; otherwise, the communication middleware would be reduced to a context middleware and no longer be compatible with non-context-aware applications. SilboPS, our implementation of content-based publish/subscribe inspired by SIENA [11, 9], solves this problem by adding two new elements to the paradigm: the context and the context function. Context represents the actual contextual information specific to the message to be sent or that needs to be notified to the subscriber, whereas the context function is evaluated using the publisher’s context and the subscriber’s context to decide whether the current message and context are useful for the subscriber. In this manner, context logic management is decoupled from context management, increasing the flexibility of communication and usage across different applications. Since the default context is empty, context-aware and classical applications can use the same SilboPS, resolving the syntactic mismatch that there is between the two categories. In any case, the possible semantic mismatch is still present because it depends on how each application interprets the data, and it cannot be resolved by an agnostic third party. The IoT environment introduces not only context but scaling challenges too. The number of sensors, the volume of the data that they produce and the number of applications that could be interested in harvesting such data are growing all the time. Today’s response to the above need is cloud computing. However, cloud computing applications need to be able to scale elastically [22]. Unfortunately there is no slicing, as distributed system primitives that support internal state partitioning [33] and hot swapping and current cloud systems like OpenStack or OpenNebula do not provide elastic monitoring out of the box. This means there is a two-sided problem: 1) how to scale an application elastically and 2) how to monitor the application and know when it should scale in or out. E-SilboPS is the elastic version of SilboPS. I t is the solution for the monitoring problem thanks to its content-based publish/subscribe nature and, unlike other solutions [5], it scales efficiently so as to meet workload demand without overprovisioning or underprovisioning. Additionally, it is based on a newly designed algorithm that shows how to add elasticity in an application with different state constraints: stateless, isolated stateful with external coordination and shared stateful with general coordination. Its evaluation shows that it is able to achieve remarkable speedups where the network layer is the main limiting factor: the calculated efficiency (see Figure 5.8) shows how each configuration performs with respect to adjacent configurations. This provides insight into the actual trending of the whole system in order to predict if the next configuration would offset its cost against the resulting gain in notification throughput. Particular attention has been paid to the evaluation of same-cost deployments in order to find out which one is the best for the given workload demand. Finally, the overhead introduced by the different configurations has been estimated to identify the primary limiting factor for throughput. This helps to determine the intrinsic sequential part and base overhead [26] of an optimal versus a suboptimal deployment. Depending on the type of workload, this can be as low as 10% in a local optimum or as high as 60% when an overprovisioned configuration is deployed for a given workload demand. This Karp-Flatt metric estimation is important for system management because it indicates the direction (scale in or out) in which the deployment has to be changed in order to improve its performance instead of simply using a scale-out policy.

Design and Implementation of a Cloud Infrastructure for Distributed Scientific Calculation

Cloud computing enables independent end users and applications to share data and pooled resources, possibly located in geographically distributed Data Centers, in a fully transparent way. This need is particularly felt by scientific applications to exploit distributed resources in efficient and scalable way for the processing of big amount of data. This paper proposes an open so- lution to deploy a Platform as a service (PaaS) over a set of multi- site data centers by applying open source virtualization tools to facilitate operation among virtual machines while optimizing the usage of distributed resources. An experimental testbed is set up in Openstack environment to obtain evaluations with different types of TCP sample connections to demonstrate the functionality of the proposed solution and to obtain throughput measurements in relation to relevant design parameters.