Network management is becoming more complex due to various factors. The growth of IoT increases the number of nodes to control. The combination of Edge and Fog Computing with distributed algorithms makes network synchronization challenging. Softwarized technologies simplify network management but create integration issues with legacy networks. Even in industrial settings where drones and mobile robots are used, proper network management is crucial yet challenging. In this context, digital twins can be used to replicate the structure and behavior of the physical network and at the same time can be used to successfully manage the complexity and heterogeneity of current networks. Despite the rapid growth of interest in the topic, a comprehensive overview of Network Digital Twin research is currently missing. To address this gap, in this paper, we present a systematic review of the Network Digital Twin literature. From the analysis of 138 primary studies, various insights emerge. Networking Digital Twin is a particularly recent concept that has been explored in the literature since 2017 and is experiencing a steady increase to this day. The vast majority of the studies propose solutions to optimize network performance, but there are also many oriented towards other goals such as security and functional suitability. The three most recurrent application domains, as self-reported in the primary studies, are those of smart industry, edge computing, and vehicular. The main research topics aim at network optimization, support for offloading, resource allocation, and floor monitoring, but also support in the implementation of machine learning algorithms such as federated learning. As a conclusion, Networking Digital Twin proves to be a promising emerging field both for academics and practitioners.

Background: Over the past decade, microservices have surged in popularity within software engineering. From a research viewpoint, mining studies are frequently employed to assess the evolution of diverse microservice properties. Despite the growing need, a validated static method to swiftly identify microservices seems to be currently missing in the literature. Aims: We present Claim, a lightweight static approach that analyzes configuration files to identify microservices in Dockerized environments, specifically designed with mining studies in mind. Method: To validate Claim, we conduct an empirical experiment comprising 20 repositories, 160 microservices, and 13k commits. A priori and manually defined ground truths are used to evaluate Claim’s microservice identification effectiveness and efficiency. Results: Claim detects microservices with an accuracy of 82.0%, reports a median execution time of 61ms per commit, and requires in the worst case scenario 125.5s to analyze the history of a repository comprising 1509 commits. With respect to its closest competitor, CLAIM shines most in terms of false positive reduction (-40%). Conclusions: While not able to reconstruct a microservice architecture in its entirety, Claim is an effective and efficient option to swiftly identify microservices in Dockerized environments, and seems especially fitted for software evolution mining studies.

Software rejuvenation is a proactive maintenance technique that counteracts software aging by restarting a system, making selection of rejuvenation times critical to improve reliability without incurring excessive downtime costs. Various stochastic models of Software Aging and Rejuvenation (SAR) have been developed, mostly having an underlying stochastic process in the class of Continuous Time Markov Chains (CTMCs), Semi-Markov Processes (SMPs), and Markov Regenerative Processes (MRGPs) under the enabling restriction, requiring that at most one general (GEN), i.e., non-Exponential, timer be enabled in each state. We present a SAR model with an underlying MRGP under the bounded regeneration restriction, allowing for multiple GEN timers to be concurrently enabled in each state. This expressivity gain not only supports more accurate fitting of duration distributions from observed statistics, but also enables the definition of mixed rejuvenation strategies combining time-based and inspection-based policies, where the time to the next inspection or rejuvenation depends on the outcomes of diagnostic tests. Experimental results show that replacing GEN timers with Exponential timers with the same mean (to satisfy the enabling restriction) yields inaccurate rejuvenation policies, and that mixed rejuvenation outperforms time-based rejuvenation in maximizing reliability, though at the cost of an acceptable decrease in availability.