An Operational Framework for Evaluating the Performance of Learning Record Stores

Nowadays, Learning Record Stores (LRS) are increasingly used within digital learning systems to store learning experiences. Multiple LRS software have made their appearance in the market. These systems provide the same basic functional features including receiving, storing and retrieving learning records. Further, some of them may offer varying features like visualization functions and interfacing with various external systems. However, the non-functional requirements such as scalability, response time and throughput may differ from one LRS to another. Thus, for a specific organization, choosing the appropriate LRS is of high importance, since adopting a non-optimized one in terms of non-functional requirements may lead to a loss of money, time and effort. In this paper, we focus on the performance aspect and we introduce an operational framework for analyzing the performance behaviour of LRS under a set of test scenarios. Moreover, the use of our framework provides the user with the possibility to choose the suitable strategy for sending storing requests to optimize their processing while taking into account the underlying infrastructure. A set of metrics are used to provide performance measurements at the end of each test. To validate our framework, we studied and analyzed the performances of two open source LRS including Learning Locker and Trax.