Computer systems can fail for a vast range of reasons, and handling failures is crucial to any critical computer system. Many modern computer systems are migrating to public clouds, which provides more flexible resource consumption and in many cases reduced costs, while the migration can also require system changes due to limitations in the provided cloud environment. This thesis evaluates a few methods of achieving failover when migrating a system to a public cloud, with the main goal of finding a replacement for failover mechanisms that can only be used in self-managed infrastructure. A few different failover methods are evaluated by looking into different aspects of how each method would change an existing system. Two methods using \textit{etcd} and \textit{Apache ZooKeeper} are used for experimental evaluation where failover time is measured in two simulated scenarios where the primary process terminates and a standby process needs to be promoted to the primary status. In one scenario, the primary process is not able to notify other processes in the system before terminating, and in the other scenario, the primary process can release the primary status to another instance before terminating. The etcd and ZooKeeper solutions are shown to behave quite similarly in the testing setup, while the ZooKeeper solution might be able to achieve lower failover time in low-latency environments.

This thesis aims to explore the use of a predictive scaling algorithm to scale a microservices-based system according to a predicted system load. A scalable system along with a predictive scaling algorithm is developed and tested by applying a periodic load to the system. The developed scaling algorithm is a combination of a reactive and a predictive algorithm, where the reactive algorithm is used to scale the system when no significant load changes are predicted. The results show that the periodical load is predicted by the algorithm, that the system can be scaled preemptively, and that the algorithm has room for improvement in terms of accuracy. / Detta examensarbete siktar på att utforska möjligheten att använda förutsägande skalningsalgoritmer för att skala ett microservices-baserat system enligt en förutspådd belastning på systemet. Ett skalbart system utvecklas tillsammans med en förutsägande skalningsalgoritm, och testas genom att applicera en periodisk belastning på systemet. Den utvecklade skalningsalgoritmen är en kombination av en reaktiv och förutsägande algoritm, där den reaktiva algoritmen används för att skala systemet då inga signifikanta belastningar förutspås. Resultaten visar att systemets belastning kan förutspås och att systemet kan skalas med hjälp av den förutspådda belastningen, samt att algoritmen har utrymme för förbättringar.

The explosion in growth of smart wireless devices has increased the ubiquitous presence of computational resources and location-based data. This new reality of numerous wireless devices capable of collecting, sharing, and processing information, makes possible an avenue for new enhanced applications. Multiple radio nodes with diverse functionalities can form a wireless cloud computing network (WCCN) and collaborate on executing complex applications using wireless distributed computing (WDC). Such a dynamically composed virtual cloud environment can offer services and resources hosted by individual nodes for consumption by user applications. This dissertation proposes an architectural framework for WCCNs and presents the different phases of its development, namely, development of a mathematical system model of WCCNs, simulation analysis of the performance benefits offered by WCCNs, design of decision-making mechanisms in the architecture, and development of a prototype to validate the proposed architecture. The dissertation presents a system model that captures power consumption, energy consumption, and latency experienced by computational and communication activities in a typical WCCN. In addition, it derives a stochastic model of the response time experienced by a user application when executed in a WCCN. Decision-making and resource allocation play a critical role in the proposed architecture. Two adaptive algorithms are presented, namely, a workload allocation algorithm and a task allocation - scheduling algorithm. The proposed algorithms are analyzed for power efficiency, energy efficiency, and improvement in the execution time of user applications that are achieved by workload distribution. Experimental results gathered from a software-defined radio network prototype of the proposed architecture validate the theoretical analysis and show that it is possible to achieve 80 % improvement in execution time with the help of just three nodes in the network. / Ph. D.

In today’s world, IoT systems are more and more overwhelming. All electronic devices are becoming connected. From lamps and refrigerators in smart homes, smoke detectors and cameras in monitoring systems, to scales and thermometers in healthcare systems, until phones, cars and watches in smart cities. All these connected devices generate a huge amount of data collected from the environment. To take advantage of these data, a processing phase is needed in order to extract useful information, allowing the best management of the system. Since most objects in IoT systems are resource limited, the processing step, usually performed by an artificial intelligence model, is offloaded to a more powerful machine such as the cloud server in order to benefit from its high storage and processing capacities. However, the cloud server is geographically remote from the connected device, which leads to a long communication delay and harms the effectiveness of the system. Moreover, due to the incredibly increasing number of IoT devices and therefore offloading operations, the load on the network has increased significantly. In order to benefit from the advantages of cloud based AIoT systems, we seek to minimize its shortcomings. In this thesis, we design a distributed architecture that allows combining these three domains while reducing latency and bandwidth consumption as well as the IoT device’s energy and resource consumption. Experiments conducted on different cloud based AIoT systems showed that the designed architecture is capable of reducing up to 80% of the transmitted data. / En el mundo actual, los sistemas de IoT (Internet de las cosas) son cada vez más abrumadores. Todos los dispositivos electrónicos se están conectando entre sí. Desde lámparas y refrigeradores en hogares inteligentes, detectores de humo y cámaras para sistemas de monitoreo, hasta básculas y termómetros para sistemas de atención médica, pasando por teléfonos, automóviles y relojes en ciudades inteligentes. Todos estos dispositivos conectados generan una enorme cantidad de datos recopilados del entorno. Para aprovechar estos datos, es necesario un proceso de análisis para extraer información útil que permita una gestión óptima del sistema. Dado que la mayoría de los objetos en los sistemas de IoT tienen recursos limitados, la etapa de procesamiento, generalmente realizada por un modelo de inteligencia artificial, se traslada a una máquina más potente, como el servidor en la nube, para beneficiarse de su alta capacidad de almacenamiento y procesamiento. Sin embargo, el servidor en la nube está geográficamente alejado del dispositivo conectado, lo que conduce a una larga demora en la comunicación y perjudica la eficacia del sistema. Además, debido al increíble aumento en el número de dispositivos de IoT y, por lo tanto, de las operaciones de transferencia de datos, la carga en la red ha aumentado significativamente. Con el fin de aprovechar las ventajas de los sistemas de AIoT (Inteligencia Artificial en el IoT) basados en la nube, buscamos minimizar sus desventajas. En esta tesis, hemos diseñado una arquitectura distribuida que permite combinar estos tres dominios al tiempo que reduce la latencia y el consumo de ancho de banda, así como el consumo de energía y recursos del dispositivo IoT. Los experimentos realizados en diferentes sistemas de AIoT basados en la nube mostraron que la arquitectura diseñada es capaz de reducir hasta un 80% de los datos transmitidos.

Yes / By 2020, next generation (5G) cellular networks are expected to support a 1000 fold traffic increase. To meet such traffic demands, Base Station (BS) densification through small cells are deployed. However, BSs are costly and consume over half of the cellular network energy. Meanwhile, Cloud Radio Access Networks (C-RAN) has been proposed as an energy efficient architecture that leverage cloud computing technology where baseband processing is performed in the cloud. With such an arrangement, more energy gains can be acquired through statistical multiplexing by reducing the number of BBUs used. This paper proposes a green Intelligent Traffic and Resource Elastic Energy (iTREE) scheme for C-RAN. In iTREE, BBUs are reduced by matching the right amount of baseband processing with traffic load. This is a bin packing problem where items (BS aggregate traffic) are to be packed into bins (BBUs) such that the number of bins used are minimized. Idle BBUs can then be switched off to save energy. Simulation results show that iTREE can reduce BBUs by up to 97% during off peak and 66% at peak times with RAN power reductions of up to 27% and 18% respectively compared with conventional deployments.

Cloud computing has become a powerful enabler for many IT services and new technolo-gies. It provides access to an unprecedented amount of resources in a fine-grained andon-demand manner. To deliver such a service, cloud providers should be able to efficientlyand reliably manage their available resources. This becomes a challenge for the manage-ment system as it should handle a large number of heterogeneous resources under diverseworkloads with fluctuations. In addition, it should also satisfy multiple operational require-ments and management objectives in large scale data centers.Autonomic computing techniques can be used to tackle cloud resource managementproblems. An autonomic system comprises of a number of autonomic elements, which arecapable of automatically organizing and managing themselves rather than being managedby external controllers. Therefore, they are well suited for decentralized control, as theydo not rely on a centrally managed state. A decentralized autonomic system benefits fromparallelization of control, faster decisions and better scalability. They are also more reliableas a failure of one will not affect the operation of the others, while there is also a lower riskof having faulty behaviors on all the elements, all at once. All these features are essentialrequirements of an effective cloud resource management.This thesis investigates algorithms, models, and techniques to autonomously managejobs, services, and virtual resources in a cloud data center. We introduce a decentralizedresource management framework, that automates resource allocation optimization and ser-vice consolidation, reliably schedules jobs considering probabilistic failures, and dynam-icly scales and repacks services to achieve cost efficiency.As part of the framework, we introduce a decentralized scheduler that provides andmaintains durable allocations with low maintenance costs for data centers with dynamicworkloads. The scheduler assigns resources in response to virtual machine requests andmaintains the packing efficiency while taking into account migration costs, topologicalconstraints, and the risk of resource contention, as well as fluctuations of the backgroundload.We also introduce a scheduling algorithm that considers probabilistic failures as part ofthe planning for scheduling. The aim of the algorithm is to achieve an overall job reliabil-ity, in presence of correlated failures in a data center. To do so, we study the impacts ofstochastic and correlated failures on job reliability in a virtual data center. We specificallyfocus on correlated failures caused by power outages or failure of network components onjobs running large number of replicas of identical tasks.Additionally, we investigate the trade-offs between vertical and horizontal scaling. Theresult of the investigations is used to introduce a repacking technique to automatically man-age the capacity required by an elastic service. The repacking technique combines thebenefits of both scaling strategies to improve its cost-efficiency. / Datormoln har kommit att bli kraftfulla möjliggörare för många nya IT-tjänster. De ger tillgång till mycket storskaliga datorresurser på ett finkornigt och omedelbart sätt. För att tillhandahålla sådana resurser krävs att de underliggande datorcentren kan hantera sina resurser på ett tillförlitligt och effektivt sätt. Frågan hur man ska designa deras resurshanteringssystem är en stor utmaning då de ska kunna hantera mycket stora mängder heterogena resurser som i sin tur ska klara av vitt skilda typer av belastning, ofta med väldigt stora variationer över tid. Därtill ska de typiskt kunna möta en mängd olika krav och målsättningar för hur resurserna ska nyttjas. Autonomiska system kan med fördel användas för att realisera sådana system. Ett autonomt system innehåller ett antal autonoma element som automatiskt kan organisera och hantera sig själva utan stöd av externa regulatorer. Förmågan att hantera sig själva gör dem mycket lämpliga som komponenter i distribuerade system, vilka i sin tur kan bidra till snabbare beslutsprocesser, bättre skalbarhet och högre feltolerans. Denna avhandling fokuserar på algoritmer, modeller och tekniker för autonom hantering av jobb och virtuella resurser i datacenter. Vi introducerar ett decentraliserat resurshanteringssystem som automatiserar resursallokering och konsolidering, schedulerar jobb tillförlitligt med hänsyn till korrelerade fel, samt skalar resurser dynamiskt för att uppnå kostnadseffektivitet. Som en del av detta ramverk introducerar vi en decentraliserad schedulerare som allokerar resurser med hänsyn till att tagna beslut ska vara bra för lång tid och ge låga resurshanteringskostnader för datacenter med dynamisk belastning. Scheduleraren allokerar virtuella maskiner utifrån aktuell belastning och upprätthåller ett effektivt nyttjande av underliggande servrar genom att ta hänsyn till migrationskostnader, topologiska bivillkor och risk för överutnyttjande. Vi introducerar också en resursallokeringsalgoritm som tar hänsyn till korrelerade fel som ett led i planeringen. Avsikten är att kunna uppnå specificerade tillgänglighetskrav för enskilda tjänster trots uppkomst av korrelerade fel. Vi fokuserar främst på korrelerade fel som härrör från problem med elförsörjning och från felande nätverkskomponenter samt deras påverkan på jobb bestående av många identiska del-jobb. Slutligen studerar vi även hur man bäst ska kombinera horisontell och vertikal skalning av resurser. Resultatet är en process som ökar kostnadseffektivitet genom att kombinera de två metoderna och därtill emellanåt förändra fördelning av storlekar på virtuella maskiner.

Cloud Computing has started to become a viable option for computing centers and mobile consumers seeking to reduce cost overhead, power consumption, and increase software services available within their platform. For instance distributed memory constrained mobile devices can expand their ability to share real time data by utilizing virtual memory located within the cloud. Cloud memory services can be configured to restrict read and write access to the shared memory pool on a partner by partner basis. Utilization of such resources in turn reduces hardware requirements on mobile devices while lessening power consumption for each physical resource. Within the Cloud Computing paradigm, computing resources are provisioned to consumers on demand and guaranteed through service level agreements. Although the idea of a computing utility is not new, its realization has come to pass as researchers and corporate companies embark on a journey of implementing highly scalable cloud environments. As new solutions and architectures are proposed, additional use cases and consumer concerns have been revealed. These issues range from consumer security, adequate service level agreements and vendor interoperability, to cloud technology standardizations. Further, the current state of the art does not adequately address these needs for mobile consumers, where services need to be guaranteed even as consumers dynamically change locations. Due to the rapid adoption of virtualization stacks and the dramatic increase of mobile computing devices, cloud providers must be able to handle logical and physical mobility of consumers. As consumers move throughout geographical regions, there exists the probability that a consumer’s new locale may hinder a producer’s ability to uphold service level agreements. This inability is due to the fact that a producer may not have physical resources located relatively close to a mobile consumer’s new locale. As a consequence, producers must either continue to provide degraded resource consumption or migrate workloads to third party producers in order to ensure service level agreements are maintained. The goal of this report is to research existing architectures that provide the ability to adequately uphold service level agreements as mobile consumers move from locale to locale. Further we propose an architecture that can be implemented along with existing solutions in order to ensure consumers receive adequate service levels regardless of locality. We believe this architecture will lead to increased cloud interoperability and decreased consumer to producer platform coupling. / text

No description available.

No description available.

Pelningumas, tikslumas, marketingo strategija ir Ichimoku Cloud prognozavimo metodo architektūros principai yra pagrindinė tema šiame darbe. Pagrindinis tikslas yra ištirti pelningumą, taikant šį metodą, siekiant jį padidinti keičiant Ichimoku Cloud struktūrą. Pirmiausia, yra sprendžiami uždaviniai, nagrinėjami grafikai ir jų teikiama informacija, kuri parodo prekiavimo strategijas, kurios naudojamos pelningumui paskaičiuoti. Atliekamas eksperimentas ir yra palyginti modifikuoto ir standartinio atvejų rezultatai. Pasirinktas S & P500 indeksas mokslinio tyrimo duomenys ir viena iš galimų prognozavimo metodo strategijų - tenkan sen ir kijun sen indikatorių sankirtos strategija. Pasirinktas penkerių metų duomenų laikotarpis. Nubrėžiamos kitos tyrimo ribos. Tyrimo eiga prasideda indikatorių reakcijos tyrimu keičiant periodų reikšmes, vėliau rezultatai palyginami su kitais atvejais, kurie buvo minimi literatūroje. Rezultatų ir analizių grafikai suteikia daug informacijos eksperimentiniam tyrimui. Pelningumas standartiniu atveju Ichimoku Cloud pelningumas yra 45,41% (40,41% su komisiniu) su 40% tikslumu ir 0,9% vidutiniu sandorio pelningumu. Autorius pakeistu atveju pelningumas 22,81% didesnis, palyginus su standartiniu atveju ir siekia 68,22% (63,22% su komisiniu) su 42% tikslumu ir 1,36% vidutiniu pelningumu. Nauja struktūra, kuri buvo rasta moksliniu tyrimu, pasiteisino su Londono (FTSE100) ir Japonijojos (NIKKEI225) indeksai, bet neveikė su kitu JAV indeksu (Nasdaq). Ichimoku... [toliau žr. visą tekstą] / Profitability, accuracy, marketing strategy and a set of architecture principles of Ichimoku Claud forecasting method is the main theme of this work. The main goal is to explore the profitability of this method in order to increase it by changing the structure of the Ichimoku Cloud. First of all, there are solved tasks by examining the structure of a set of graphs and information provided by marketing strategies that are used in profitability analysis. Experiment is carried out and there are compared results between modified and standart cases. There are S&P500 index selected in the research and one of the possible forecasting method‘s strategies – tenkan sen and kijun sen indicators cross strategy. Five year data period is selected. There are plotted other boundaries of the research. Experimental procedure begins with the research of the reaction of the indicators by changing periods of time and calculation of profitability, later on there are results compared with other cases, which were mentioned in the literature. The results and analysis of graphs provides lots of information to do experimental research. Profitability of standart case of Ichimoku Cloud is 45,41% (40,41% with the commission) with an accuracy of 40% and 0,9% average return of the transaction. Author’s modified case yield 22,81% more profitability compared to standart case and reaches 68,22% (63,22% with the commission) with an accuracy of 42% and 1,36% average return of transaction. New structure, which... [to full text]