Paxos is an algorithm for implementing fault-tolerant distributed systems. The core of Paxos is found in many consensus algorithms. Raft and Zab are two prominent protocols that are used in the industry. They serve as the foundation of distributed key-value stores and coordination services such as Consul, Etcd and ZooKeeper. In distributed consensus, the most common way for servers to agree over a value is to use majority quorums. However in 2016, FPaxos (Flexible Paxos) was published. The authors make the observation that majority quorums are not required as intersection is mandatory only across the two phases of Paxos. By taking advantage of this, developers will have more control over the choice between performance and availability. In this paper we will look at how FPaxos can be adopted into existing systems and demonstrate the advantages through a ZooKeeper modification running on City Cloud.

No description available.

Denna rapport handlar om ett projekt åtta studenter på Linköpings universitet genomförde våren 2017, inom ramarna för kursen TDDD96: Kandidatprojekt i programvaruutveckling. Fokuset i projektet var att bygga upp en ny webbplats åt en kund, Musikaliska konstföreningen, som ska ersätta deras gamla webbplats. Projektgruppen valde att göra detta med hjälp av verktyget WordPress. Gruppen valde också att göra en alternativ webbplats med programmeringsspråket Python och ramverket Flask, för att sedan jämföra resultatet mot WordPress-resultatet. Det gruppen valde att jämföra är vilket stöd WordPress respektive Flask ger åt utvecklarna, och vilket av dem som ger lämpligast resultat för kunden. Resultatet blev en färdig hemsida uppbyggd med WordPress, och en prototypsida uppbyggd med Python och Flask. Resultatet visar att det är enklare att få en färdig sida med hjälp av WordPress, men att man genom att skriva kod från grunden i Flask kan få ett betydligt mer användaranpassat system. När det kommer till faktorer som driftkostnader och stöd så visade det sig att WordPress har en klar fördel i jämförelse med Flask.

Internet of Things (IoT) has revolutionized transportation systems by connecting vehicles consequently enabling their tracking, as well as monitoring of driver activities. The IoT platform for most vehicles typically consists of 1) an on-board system consisting of the communication unit, sensors and a set of ECU’s that are interconnected using a CAN network, 2) an off-board system consisting of the applications deployed on the servers (e.g., cloud) that processes the data send by the communication unit over the internet, and 3) mobile devices like a mobile phone or a computer that communicates with the on-board and off-board systems. Such an IoT platform requires a significant amount of data to be send from the on-board system to the off-board servers, contributing to high network usage. There are two main architectural paradigms for sending data: 1) interval based architecture, in which data is send at regular intervals and 2) event based architecture, in which data is send whenever relevant events occur. Currently, (e.g., at Scania), the data is being send at regular intervals, i.e., using an interval based approach. In this case, data is send even if it is not relevant for reporting leading to a wastage of network resources, e.g., when the data does not change considerably compared to the previously sent value. Sending data in an event-based manner, when the data is relevant for reporting, e.g., changes significantly, reduces the network usage when compared to the interval based approach.  In this thesis, we investigate the possibility of using an event based architecture to send data from the on-board system to the off-board system in order to reduce network usage and improve the accuracy of the data available off-board. We first propose an event based architecture for data transfer in the context of Internet of vehicles. We then implement a simulator to evaluate our proposed architecture for the specific case of position data. Finally, we perform extensive experiments varying different parameters and compare, for example, average message size per minute and average off-board error distance. The results show that our event based architecture improves the accuracy of data available at the off-board system, by a careful selection of events. Moreover, we found that our event based architecture significantly decreases the frequency of sending messages, particularly during highway driving, leading to reduced average data transfer rates. Our results enable a customer to perform trade-offs between accuracy and data transfer rates. Future work will aim at implementing the event based architecture on a real platform as well as investigating the possibility of using the event based architecture for more accurate prediction by incorporating additional details such as the final destination of the vehicle and odometer values.

Dependency parsing is important in contemporary speech and language processing systems. Current dependency parsers typically use the multi-class perceptron machine learning component, which classifies based on millions of sparse indicator features, making developing and maintaining these systems expensive and error-prone. This thesis aims to explore whether replacing the multi-class perceptron component with an artificial neural network component can alleviate this problem without hurting performance, in terms of accuracy and efficiency. A simple transition-based dependency parser using the artificial neural network (ANN) as the classifier is written in Python3 and the same program with the classifier replaced by a multi-class perceptron component is used as a baseline. The results show that the ANN dependency parser provides slightly better unlabeled attachment score with only the most basic atomic features, eliminating the need for complex feature engineering. However, it is about three times slower and the training time required for the ANN is significantly longer.

Data streaming nowadays is one of the most used approaches used by websites and applications to supply the end user with the latest articles and news. As a lot of news websites and companies are founded every day, such data centers must be flexible and it must be easy to introduce a new website to keep track of. The main goal of this project is to investigate two frameworks where implementing a robot for given website should take some acceptable amount of time. It is really challenging task, first of all it aims optimizing of a framework which means to put less efforts on something and have the same result and one another thing is that it will be used by professors and students at the end so quality and robustness play big role here. In order to overcome this challenge two different types of news websites were investigated and through this process the approximately time to implement a single robot was extracted. Having in mind the time spent to implement a single robot, the new frameworks were implemented with the goal to spend less time to implement a new web robot. The results are two general frameworks for two different types of websites, where implementing a robot does not take so much efforts and time. The implementation time of a new robot was reduced from 18 hours to approximately 4 hours.

No description available.

Context. Database is an essential part of any software product. With an emphasis on application performance, database efficiency becomes one of the key factors to analyze in the process of technology selection. With a development of new data models and storage technologies, the necessity for a comparison between relational and non-relational database engines is especially evident in the software engineering domain. Objectives. This thesis investigates current knowledge on database performance measurement methods, popularity of relational and non-relational database engines, defines characteristics of databases, approximates their average values and compares the performance of two selected database engines.Methods. In this study a number of research methods are used, including literature review, a review of Internet sources, and an experiment. Literature datasets used in the research incorporate over 100 sources including IEEE Xplore and ACM Digital Library. YCSB Benchmark was used as a direct performance comparison method in an experiment to compare OracleDB’s and MongoDB’s performance. Results. A list of database performance measurement methods has been defined as a result of the literature review. Two most popular database management engines, one relational and one non-relational has been identified. A set of database characteristics and a database performance comparison methodology has been identified. Performance of two selected database engines has been measured and compared. Conclusions. Performance comparison between two selected database engines indicated superior results for MongoDB under the experimental conditions. This database proved to be more efficient in terms of average operation latency and throughput for each of the measured workloads. OracleDB however, presented stable results in each of the category leaving the final choice of database to the specifics of a software engineering project. Activities required for the definition of database performance comparison methodology proved to be challenging and require study extension.

Recent years have seenincreased use of open source-based tools for development of evensecurity-critical industrial systems. Shown here is a demonstration prototypeof a control system for airport landing lights. The project is written in theC++ programming language for the open source-based HMI/SCADA platform Pvbrowserand deployed for the Ubuntu Linux operating system, and serves to illustrateone practical application and the general feasibility of open source-basedtools as a cost-efficient alternative within industrial control systems development. / På senare tid har användningen av verktyg baserade påöppen källkod ökat, ävenför utveckling av säkerhetskritiska industriellasystem. Här redovisas endemonstrationsprototyp till ett styrsystem förlandningsljus på flygplats. Projektet är skrivet i programspråket C++ för denopen source-baserade HMI/SCADA-plattformen Pvbrowser och driftsatt föroperativsystemet Ubuntu Linux, och illustrerar ett exempel på praktisktillämpning för samt rimligheten i att använda verktyg baserade på öppenkällkod som ett kostnadseffektivt alternativ för utveckling av industriellastyrsystem.

Denna rapport beskriver det examensarbete som genomförts av Johan Möller. Examensarbetets syfte var att utveckla en lösning för att flytta virtuella maskiner mellan olika datastores som finns i virtuella miljöer. Examensarbetet utfördes på Atea Sverige AB som hyr ut virtuella serverlösningar till sina kunder. Arbetet är utfört på 10 veckor, där den största delen av tiden har gått åt till att sätta upp en lämplig virtuell testmiljö som liknar Ateas riktiga produktionsmiljö. Av rapporten framgår hur jag har satt upp den virtuella testmiljön. Jag redovisar även utvecklingsarbetet som genomförts för att få mitt script för att flytta virtuella maskiner mellan olika datastores att fungera på det sätt jag ville, det vill säga utföra uppgiften på ett så enkelt och säkert sätt som möjligt. I rapporten tas det också upp hur lämplig VMwares vCloud-svit är för automation i virtuella miljöer. Min slutsats av examensarbetet är att VMwares produkter håller en hög  standard med mängder av funktioner, dock med en del buggar som kan vara irriterande för användaren. Jag har fått en bra insyn i hur virtuella miljöer fungerar i så här stor skala och vilka verktyg som finns för att administrera över dem.