In this paper we perform performance measurements of six different servers that can function as key-value stores, ranging from purpose made key-value store memcached, over more complex Redis to MySQL, a traditional relational database that also implements access by memcached protocol. We analyze the influence of number of simultaneous clients and item data size on measured wall time of client and server components.

This article presents a survey of model-driven techniquesfor data model synthesis. During an extensive research,we identified more than 70 research papers in the field andmore than 15 different graphical notations used for the sourcemodel representation.We have classified the proposed approachesinto four distinct groups: function-oriented, process-oriented,communication-oriented and goal-oriented. Their contributionsare presented in chronological order and evaluated based onseveral main criteria. Although the idea of model-driven designof the data model is more than 25 years old, the survey showsthe richness and diversity of ideas, but only a small number ofimplemented automatic generators.

This paper presents an approach to the automated design of the initial conceptual database model. The UML activity diagram, as a frequently used business process modeling notation, is used as the starting point for the automated generation of the UML class diagram representing the conceptual database model. Formal rules for automated generation cover the automatic extraction of business objects and business process participants, as well as the automatic generation of corresponding classes and their associations. Based on these rules we have implemented an automatic generator and evaluated it on a real business model.

In this paper we have analysed the impact of running transactions at different isolation levels on application performance, more specifically on the execution time of the set of transactions. Three scenarios of transaction executions were defined, based on real applicative cases, and applied to different transaction sets and transaction isolation levels. Tests were performed using three actual DBMS: Oracle, Microsoft SQL Server and MySQL.