Entwurf einer einheitlichen API zur Nutzung von Mellanox V-API und OpenIB Verbs auf Basis von C Pre-Prozessor Makros und Integration der API in das vorhandene MPICH2-CH3 Device für Infiniband

Įmonė UAB „Duomenų Centras“ nuomoja serverius jau 13 metų. Atlikus tyrimus paaiškėjo, kad įmonė eksportuoja apie 75% savo paslaugų į daugiau nei 130 pasaulio šalių. Prie šio rezultato labai prisidėjo Duomenų Centro partneriai iš Lietuvos ir kitų šalių. Partneris – tai svarbiausias klientas, kuris pritraukia kitus naujus klientus. Per įmonės gyvavimo metus visas perpardavinėjimo ir partnerių klientų paslaugų valdymo procesas buvo per elektorinį paštą - tai užtrunka, tokios paslaugos nėra itin patrauklios šiuolaikinėje interneto rinkoje. Todėl dokumente atskleidžiama „partnerių programa“, bei API posistemė, kuri leidžia įmonės partneriams vykdyti su paslaugomis susijusias funkcijas. / The modern data center of Balticservers have been selling their services now for 13 years. The research revealed that the company exports 75% of its services to more than 130 countries worldwide. Moreover, the company has partners from Lithuania and many other countries. The product has been designed for partners that optimizes the process of reselling Balticservers services. It offers a WHMCS module which allows an easy creation and customization of services that are desired to be resold from Balticservers. Also, this includes an API communication between two separated systems that are WHMCS and Korys which Balticservers are currently developing and using in their daily work. API has been developed by using PHP scripting language, cURL and JSON technologies. The document exposes a few similar API systems’ comparison, product requirement specification, architecture’s specification, testing protocol and partners API documentation.

Submitted by Kilvya Braga (kilvyabraga@hotmail.com) on 2018-05-17T13:52:52Z No. of bitstreams: 1 FILIPE NEVES CAVALCANTE - DISSERTAÇÃO (PPGCC) 2017.pdf: 1391449 bytes, checksum: bf8690a386cab833342c9beaf39b1211 (MD5) / Made available in DSpace on 2018-05-17T13:52:52Z (GMT). No. of bitstreams: 1 FILIPE NEVES CAVALCANTE - DISSERTAÇÃO (PPGCC) 2017.pdf: 1391449 bytes, checksum: bf8690a386cab833342c9beaf39b1211 (MD5) Previous issue date: 2017 / Java Collections Framework (JCF) oferece um conjunto de Tipos Abstratos de Dados (TADs) na linguagem Java, sendo um framework bastante utilizado. Entretanto, as particularidades de uso de cada coleção são negligenciadas em favor da praticidade, sendo muito comum, por exemplo, o uso de listas sem permitir repetições de elementos, como se fossem um TAD conjunto. Há abordagens semiautomáticas de correção desse tipo de problema, mas elas introduzem classes adaptadores ao código original, o que nem sempre é desejável. Já a Substituição Direta apenas troca uma chamada de método por outra(s) correspondente(s), realizando o mínimo de mudanças possíveis, parecendo ser uma opção melhor. Entretanto, ela é mais complexa, o que leva este trabalho a estudar as condições e consequências para que ela seja possível. Foram desenvolvidas transformações implementadas num ambiente de reescrita de programas, sendo estas testadas e analisadas por meio de um estudo de caso, a fim de verificar sua aplicabilidade, e também através de questionário destinado a programadores Java, que analisou a aceitabilidade das transformações em dois trechos de código. Como resultados do estudo de caso,tem-se a ocorrência de14% de projetos fazendo uso de listas atuando como conjuntos. E a ArrayList apresentou maior incidência (90%) já List ocorreu nos demais casos (10%). Em se tratando das avaliações realizadas pelos desenvolvedores Java no questionário, cerca de 45% dos participantes propuseram sugestões consideradas adequadas, de acordo com a API e JCF [1], para um trecho de código que utilizava o método add(E e) de List, e 25% deles para outro trecho que realizava uma chamada do método set(int index, E e). Com isso, percebe-se a relevância do problema, sendo necessário continuar desenvolvendo novas técnicas e ferramentas que façam um programa utilizar coleções de forma adequada, tornando-o mais legível e eficiente. / Java Collections Framework (JCF) offers a set of Abstract Data Types (ADTs) in the Java language, being a widely used framework. However, because particularities of each collection the usage are neglected in favor of practicality, beingvery common,for example, the use of lists without allowing repetitions of elements, as if they were a set ADT. There are semiautomatic approaches to correct this type of problem, but they insert adaptive classes to the original code, which is not always desirable. Direct substitution, however, only exchanges a method call for another(s) corresponding(s), making as few changes as possible, appearing to be a better option. However, it is more complex ,which leads this work to study the conditions and consequences to make it possible. It was developed transformations implemented in a program rewriting environment, which were tested and analyzed through a case study, anend to verify its application, and also through a questionnaire aimed at Java programmers, which analyzed the acceptability of transformations in two sections of code. Asacasestudy results, there is an occurrence of 14% of the works that make use of lists acting as sets. And an ArrayList presented higher incidence (90%) and List itself cases (10%). Regarding the revisions made by the Java developers in the questionnaire, about 45% of the participants proposed suggestions considered appropriate, according to an API and JCF [1], for a section of code that used the add (E e) method of List , And 25% of them for another section that performed a call of the method set (int index, E and). With this, we can see the relevance of the problem, being necessary to continue to develop new techniques and tools that will make a program using collection properly, making it more readable and efficient .

O objetivo deste trabalho é avaliar a recuperação de imagem por conteúdo (CBIR - Content-based Image Retrieval) em uma rede de computadores, estabelecendo-se métricas de controle que otimizem a utilização da rede e ao mesmo tempo garanta melhor qualidade na resposta. Prevê-se a recuperação de imagens baseada em consulta por similaridade, combinando-se o extrator de características com a função de distância. O desenvolvimento é realizado em uma linguagem de programação independente da plataforma e de âmbito de internet, utilizado Java, desenvolvendo-se uma API (Application Programming Interface), visando, em especial, a reutilização de código, o que implica na diminuição do tempo de desenvolvimento. O trabalho demonstra a utilização de estruturas de indexação sequencial, comparada com a estrutura de indexação de árvore \"Slim-tree\". A principal contribuição deste trabalho é a análise e a utilização de métricas na rede de computadores em um projeto de CBIR. / The objective of this study is to evaluate the Content-based Image Retrieval (CBIR) in a computer network, establishing control metrics that optimize the use of network while ensuring better quality in the response. It is expected that images are going to be recovered based on queries that are similar, combining the features extractor with the distance function. The development environment is an independent language of the platform and of the Internet scope, using Java, that has developed an Application Programming Interfaces (API), aiming primarily the code reuse, which implies a decrease in the development time. This thesis demonstrates the use of sequential indexing structures compared to the tree index structure \"Slim-tree\". The main contribution of this thesis is the analysis and utilization of metrics in a computer network project in CBIR.

The purpose of this thesis was to come up with a new architecture for an existing web based planning and administration tool at Ericsson. The main objective was to create a new architecture with a proof of concept implementation. To find out how the existing tool were being used we interviewed people with different roles within the company. We made observations as they performed their day-to-day activities in the tool. We then designed and implemented a new version of the tool. We came up with a database design that handles a very dynamic use of data types. On top of the database we created REST API to allow data to be easily consumed by a web interface and other internal systems at Ericsson. We also showed that the architecture we ended up with performed well even with a lot of objects in the database. We did also come up with a solution to allow people to work simultaneously with the same parts of the system without overwriting each other’s changes.

As a result of Web 2.0 technologies such as Asynchronous JavaScript and XML (Ajax) web-based applications with rich contents are evolving to be more and more like normal applications in aspects, such as interactivity, functionality, and usability. This evolvement makes it possible to create web-based services, providing maps for users to search and browse geographic information. This thesis is an evaluation of functionality, usability and accuracy for the four web-based map APIs: Google Maps, Microsoft Virtual Earth, Multimap and ViaMichelin. The thesis explains how web-based mapping works, common functionality provided, and evaluates the functionality provided by each map service provider as well as the offered usability. In addition to this, it also includes the results of several tests, illustrating the APIs’ browser compatibility, performance and accuracy. After testing and evaluation of the four APIs, the conclusion is that none of them can be appointed as the winner. They all have benefits and drawbacks; differences in terms of functionality, compatibility, usability, geocoding and development support, and the choice of API is consequently dependent of the type of application. As a result of this, and the fact that the APIs are constantly changing in terms of functionality and coverage, it is important to create applications independent of the map service provider. This was successfully done during the internship at Amadeus by creating a map abstraction layer in-between the applications and the maps, creating the possibility to switch API, or map service provider, without changed the code.

I denna rapport så redogör vi för vårt arbete med att ta fram en prototyp till bokföretaget Publit. Vi fick i uppgift av Publit att undersöka möjligheterna för dem att med hjälp av sina egna API:er och API:erna från två andra företag i samma bransch, Valobox och Jellybooks, ta fram en ny produkt eller tjänst. Vi redogör för de olika faserna som vi har gått igenom och de beslut vi tagit angående grafisk design, interaktion och kodning. Resultatet av projektet blev en fungerande prototyp som visar hur Publit med hjälp av Valobox kan skapa en ny funktion till sin widgetshop som gör så att användarna kan provläsa böcker innan de köper dem. / In this report we describe our work with producing a prototype for Publit, which is a book-company. Our assignment from Publit was to explore the possibilities when working with their API and the APIs from two other companies in the same line of business, Valobox and Jellybooks, and by doing this develop a new product or service. We present the phases which we encountered and which decisions we have made regarding graphical design, interaction and coding. The final result of the project is a working prototype which shows how Publit, with the help of Valobox, can create a new function for their widget shop. The function allows the users to sample read a book before they decide whether they want to buy it or not.

Rapporten grundar sig på en delmängd av ett projekt som utfördes åt uppdragsgivaren SocialWorks. Projektet innefattade bland annat att arbeta mot externa webbtjänster. Genom webbtjänster kan webbaserade datorprogram kommunicera och skicka data till varandra. Då en webbtjänst nyttjas behöver utvecklaren sätta sig in i API:et för webbtjänsten. Denna rapport tar upp ett antal problem som kan uppstå vid användandet av en sådan webbtjänst och hur dessa problem kan lösas genom att skapa ett mellanlager i form av en serviceklass. Alla API-anrop ska gå genom serviceklassen och därmed behöver kodförändringar endast göras i serviceklassen ifall API:et förändras. Som ett resultat av arbetet togs en serviceklass fram som underlättar användningen av Facebooks Graph API tillsammans med PHP-ramverket Zend Framework. Genom serviceklassen behöver användaren inte sätta sig in i Facebook Graph API i detalj. Dessutom skapades en testapplikation i Zend dels för att testa serviceklassens funktionalitet och dels för att hantera så kallade testanvändare som Facebooks Graph API tillhandahåller. / This report is based on a subset of a project that was made for the company SocialWorks. The project included working with web services. Through web services, web applications are able to communicate and send data to each other. When a web service is used, the user has to study the web service API. This report addresses a number of problems that might arise from the use of such web services and how these problems can be solved by creating a service class that is used as a middle layer. All API calls will be made through the service class and thus changes in the API will only require modification of code in the service class. As a result of this report a service class was created using the PHP framework Zend Framework to facilitate the use of Facebook's Graph API. By using the service class, the user does not have to study the Facebook Graph API in detail. A test application was created to test the service class functionality and manage test users provided by the Facebook Graph API.

This is a survey of different techniques that may be used for embedding geographicdata into web pages. Different models of GIS communication are examined and a testsubject in the form of an application is implemented and evaluated.The test application (‘StudentKarta’ eng. ‘Student Map’) lets students submit theirresidential data and shows their position as pins on an embedded Map.An expire date is used to let the pins fade away and disappear.The application uses the popular Google map engines to display the geographicalpresentations. Both 2D (Google Maps) and 3D (Google Earth) are used and the pinsare rendered by using bitmaps and 3-models. / Detta arbete undersöker de faktorer som kan vara avgörande vid val av teknik för attbädda in geografisk information i webbsidor varefter verktyg väljs och en pilotapplikationimplementeras och utvärderas.Som leverantör av geografiska data valdes den populära leverantören Google, förutveckling av applikationen ’StudentKarta’.Applikationen låter studenter mata in sina uppgifter varefter deras bostadsortpresenteras på en inbäddad karta. Uppgifterna är försedda med ett bäst-före-datumoch studentmarkörerna ”tonar bort” ju närmre detta datum kommer, för att slutligenförsvinna.Den resulterande applikationen presenterar en inbäddad lösning med två olikapresentationstekniker:- En tvådimensionell kartrepresentation med markörer som renderas medrastergrafik.- En tredimensionell ”Glob” (Google Earth) där positionerna märks ut medtredimensionella markörer.

The networking of control devices is well advanced in a modern automobile. To ensure secure communications with each other and to prevent mistakes due to communication error, protocol tests are performed at the physical and data link layer level. In a modern car the diagnosis of the electronic devices (e.g. reading/deleting fault memory entries) is done with a standardized communication protocol (UDS - Unified Diagnostic Services). Modern control units have this diagnostic capabilities to detect this communication error among others. This data is provided for a guided troubleshooting the vehicle via a defined interface. In this communication test detection mechanisms are specifically stimulated and read afterwards. In order to read or delete these entries, a diagnosis query is performed on the basis of the diagnostic protocol UDS. The diagnostic requests are currently performed with a diagnostic service of bus analysis software Vector CANoe. This service requests can be performed directly on CAN but for the diagnosis of other bus systems a software gateway is used. The maintenance required and the error rate of the software gateways is very high and should be replaced in the future. Furthermore, additional bus systems can be integrated only with great effort. Therfore the main aim of the master thesis was to develop a concept for univesal UDS API and afterwards implementation of that concept using CAPL programming for modular use in test environments for vehicle communication testing.