This thesis is focused on optimization of data reading from distributed NoSQL database Apache HBase with regards to the desired data granularity. The assignment was created as a product request from Seznam.cz, a.s. the Reklama division, Sklik.cz cost center to improve user experience by making filtering of aggregated statistical data available to advertiser web application users for the purpose of viewing entity performance history.

Increasing performance of computers and ability to connect computers with high speed communication networks make distributed databases systems an attractive research area. In this study, we evaluate communication and data processing capabilities of a HPC machine. We calculate accurate cost formulas for high volume data communication between processing nodes and experimentally measure sorting times. A left deep query plan executer has been implemented and experimentally used for executing plans generated by two different genetic algorithms for a distributed database environment using message passing paradigm to prove that a parallel system can provide scalable performance by increasing the number of nodes used for storing database relations and processing nodes. We compare the performance of plans generated by genetic algorithms with optimal plans generated by exhaustive search algorithm. Our results have verified that optimal plans are better than those of genetic algorithms, as expected.

Thesis discusses distributed database systems and its advantages and disadvantages. Further, text informs about document database CouchDB, storage of documents, synchronization and CAP theorem. The aim of the thesis is to implement distributed document management system and workflow management system based on CouchDB. The system contains a cluster with three CouchDB nodes with HAProxy in front, which does load balancing. The system allows creation of any document based on the template, manages its life cycle and workflow. It is also possible to create a custom workflow using BRMS rules. The implemented solution simplifies document management and workflow and allows a high degree of customization for the organizations needs.

This thesis is focused on the distributed Big Data processing on the Java platform, together with graph databases. It analyses several graph database distributions and the possibilities to connect them to the Apache Hadoop system for distributed data processing. For the purpose of testing database solutions effectiveness, the thesis outcome is an application, which is downloading data from social networks Twitter and Facebook. It is able to write and analyse data with two different database frameworks which are Halyard and HGraphDB.

No description available.

Depuis de nombreuses années, plusieurs communautés telles que IMS (Intelligent Manufacturing Systems), HMS (Holonic Manufacturing System) ont suggéré l'utilisation de produits intelligents pour rendre les systèmes adaptables et adaptatifs et ont montré les bénéfices pouvant être réalisés, tant au niveau économique, qu'au niveau de la traçabilité des produits, qu'au niveau du partage des informations ou encore de l'optimisation des procédés de fabrication. Cependant, un grand nombre de questions restent ouvertes comme la collecte des informations liées au produit, leur stockage à travers la chaîne logistique, ou encore la dissémination et la gestion de ces informations tout au long de leur cycle de vie. La contribution de cette thèse consiste en la définition d'un cadre de dissémination des informations relatives au produit durant l'ensemble de son cycle de vie. Ce cadre de dissémination est associé à un nouveau paradigme qui change radicalement la manière de voir le produit et la matière. Ce nouveau concept consiste à donner la faculté au produit d'être intrinséquement et intégralement "communicant". Le cadre de dissémination des informations proposé offre la possibilité à l'utilisateur d'embarquer des informations sensibles au contexte d'utilisation du produit communicant. Outre la définition du processus de dissémination des informations, cette thèse offre un aperçu des champs de recherche, tant scientifiques que technologiques, à investiguer à l'avenir concernant le concept de "matière communicante" / Over the last decade, communities involved with intelligent-manufacturing systems (IMS - Intelligent Manufacturing Systems, HMS - Holonic Manufacturing System) have demonstrated that systems that integrate intelligent products can be more efficient, flexible and adaptable. Intelligent products may prove to be beneficial economically, to deal with product traceability and information sharing along the product lifecycle. Nevertheless, there are still some open questions such as the specification of what information should be gathered, stored and distributed and how it should be managed during the lifecycle of the product. The contribution of this thesis is to define a process for disseminating information related to the product over its lifecycle. This process is combined with a new paradigm, which changes drastically the way we view the material. This concept aims to give the ability for the material to be intrinsically and wholly "communicating". The data dissemination process allow users to store context-sensitive information on communicating product. In addition to the data dissemination process, this thesis gives insight into the technological and scientific research fields inherent to the concept of "communicating material", which remain to be explored

本研究之主要目標在於建構一個能評估分散式關聯資料庫中之特色的需求導向績效評估方法。在過去的績效評估研究中，已經有許多人對於關聯式資料庫績效評估做了多方面的努力。但是，過去的關聯式資料庫資效評估方法如：Wisconsin、AS3AP、TPC系列的Benchmarks都有著一些限制及不足的地方。 過去的關聯式資料庫績效評估方法並無法完全的評估出分散式資料庫的特殊需求及其表現。所以本研究嘗試要建立出一個能專門適用於分散式資料庫導向的績效評估方法。為了要作出此績效評估方法，本研究採用了工作量模式的研究方法。先建出分散式資料庫績效評估的工作量模式，再以其來實作出績效評估方法。工作量模式分成三部分：資料模式、交易模式、控制模式。 / This thesis is intended to design a requirements-centric database benchmark, which can evaluate the general performance of the distributed relational database systems. In the past, there are many relational database benchmarks. But the relational database benchmarks like Wisconsin, AS3AP, TPC, TP1 have some constraints. In this study, we aim to design a general-purpose distributed database workload model and implement it. To design this benchmark, we need to build our workload model. The workload model consists of three components：data model, transaction model, control model. Each model has the requirement specification language to accommodate user's workloads.

歷年發生的大型天然災害中，行動通訊系統常常會因為道路、橋樑、電力的損毀而導致嚴重癱瘓，進而影響災後救援工作的進行，行動通訊系統其實是不可靠且極為脆弱的。為使災區能夠快速地恢復通訊，本研究提出一種應急通訊系統，利用無線通訊，將這些無連線能力的基地台連接起來，並利用存活可連至核心通訊網路的基地台，建構成為一個臨時性的通訊系統，稱為應急蜂巢式行動通訊網路(Contingency Cellular Network，CCN)，供使用者在災區內進行通訊。由於資料庫的可靠度在行動通訊網路中具有舉足輕重之地位，本論文旨在利用分散式架構提高CCN網路資料庫的可靠度，進而提高系統可靠度。我們先分析資料表的特性，並依據各資料表的特性以及資源之多寡，設計相應的分散式架構。 我們利用樹狀拓樸的簡單特性，提出三種分散式架構：(1)用於群組資料表的階層備援式架構、(2)用於HLR資料表的階層備援式架構，以及(3)用於HLR資料表的鄰近階層備援式架構。這三種架構都利用了樹狀拓樸的簡單特性，大幅簡化了資料庫的查詢及更新程序。我們以可用度及平均總成本，以評估各架構在不同的情形下之最合適架構，並觀察在各架構下，其系統可用度與平均總成本之間的關係，由結果可知，在群組資料表儲存架構中，若追求高系統可用度，可選擇高儲存密度之架構，反之，若追求較低的成本，可選擇低儲存密度之架構。而在HLR資料表儲存架構中，階層備援式的系統可用度與平均總成本比鄰近階層備援式來得高，表示階層備援式耗費較多成本，但可用度也較高，故在建置時可依實際情況來選擇合適的架構儲存。 最後，為驗證分散式資料庫架構之可行性，我們利用數部筆記型電腦及Android平台手機來建置階層備援式及鄰近階層備援式架構，並測量在此二種架構下於通話建立時所耗費的時間，從模擬結果可知，此二種架構皆可快速建置並可立即通訊，惟鄰近階層備援式所耗費的通話建立時間略長，但仍在可容許範圍之內。 / When stricken by a catastrophic natural disaster, the efficiency of disaster response operation is very critical to life saving. The efficiency of disaster response operation is greatly depending on communication systems. However, they were usually not dependable, including cellular networks, and often crashed due to power outage and backhaul link breakage. The failure of communication systems caused a big coordination problem to many disaster response operations. This thesis proposes a Contingency Cellular Network (CCN) by connecting isolated base stations to survival base stations using long-range wireless links to recover part of cellular network functionality. People can use their own cell phones for emergency communication in the disaster areas. CCN will be able to support a large number of disaster responders with limited resources in the early hours of disasters, thus to save many lives. Because the reliability of database plays a very important role in the CCN, this thesis proposes a distributed database architecture to improve the reliability of database so as the reliability of CCN. We take advantage of the simplicity of CCN tree topology to design three distributed database architectures: (1) Hierarchical Redundancy Architecture for “Group_Member ” table, (2) Hierarchical Redundancy Architecture for “HLR” table, and (3) Hierarchical Neighboring Redundancy Architecture for “HLR” table. The tree topology greatly simplified the query and update procedures. We use availability and average total cost to analytically evaluate all three architectures trying to identify the most appropriate architecture under different circumstances. Based on our evaluation results, CCN operators can choose the most appropriate architecture according to their realistic circumstances. Finally, in order to verify the feasibility of the proposed architectures, we implemented a simplified prototype using several laptops and Android mobile phones. The Hierarchical Redundancy Architecture and the Neighboring Redundancy Architecture were implemented. The results show that both architectures can be functional with reasonable performance except that Hierarchical Neighboring Redundancy Architecture may take longer time, but in a tolerable range, to setup a call.

No description available.

With the shifting landscape of computing hardware architectures and the emergence of new computing environments (e.g., large main-memory systems, hundreds of CPUs, distributed and virtualized cloud-based resources), state-of-the-art designs of transaction processing systems that rely on conventional wisdom suffer from lost performance optimization opportunities. This dissertation challenges conventional wisdom to rethink the design and implementation of transaction processing systems for modern computing environments.<div><br></div><div>We start by tackling the vertical hardware scaling challenge, and propose a deterministic approach to transaction processing on emerging multi-sockets, many-core, shared memory architecture to harness its unprecedented available parallelism. Our proposed priority-based queue-oriented transaction processing architecture eliminates the transaction contention footprint and uses speculative execution to improve the throughput of centralized deterministic transaction processing systems. We build QueCC and demonstrate up to two orders of magnitude better performance over the state-of-the-art.<br></div><div><br></div><div>We further tackle the horizontal scaling challenge and propose a distributed queue-oriented transaction processing engine that relies on queue-oriented communication to eliminate the traditional overhead of commitment protocols for multi-partition transactions. We build Q-Store, and demonstrate up to 22x improvement in system throughput over the state-of-the-art deterministic transaction processing systems.<br></div><div><br></div><div>Finally, we propose a generalized framework for designing distributed and replicated deterministic transaction processing systems. We introduce the concept of speculative replication to hide the latency overhead of replication. We prototype the speculative replication protocol in QR-Store and perform an extensive experimental evaluation using standard benchmarks. We show that QR-Store can achieve a throughput of 1.9 million replicated transactions per second in under 200 milliseconds and a replication overhead of 8%-25%compared to non-replicated configurations.<br></div>