The emergence of the Internet of Things (IoT) introduced new classes of applications whose latency and bandwidth requirements could not be satisfied by the traditional Cloud Computing model. Consequently, the IT community promoted the cooperation of two paradigms, Cloud Computing and Edge Computing, combining large-scale computing power and real-time processing capabilities. A significant management challenge in such complex infrastructure concerns the development of efficient maintenance strategies to preserve the environment’s performance and security. While the abundant resources from the academic literature could support the design of novel maintenance solutions, extracting actionable insights from the existing approaches is challenging, given the massive number of published papers. Furthermore, existing review papers, which could help summarize the state-of-the-art, scope their investigations to the maintenance of certain components in particular scenarios. This work fills this gap with a broader literature analysis that covers maintenance strategies targeting physical and logical components in cloud, edge, and IoT environments. First, we introduce a taxonomy that organizes existing solutions according to several characteristics. Then, we review the literature following the taxonomy structure to facilitate the understanding of the research landscape and the comparison between existing works. Finally, we shed light on open challenges that represent promising research directions.

物联网 (IoT) 的出现引入了新型应用程序，传统云计算模型无法满足其延迟和带宽要求。因此，IT界推动了云计算和边缘计算两种范式的合作，将大规模计算能力和实时处理能力结合起来。如此复杂的基础设施中的一个重大管理挑战涉及制定有效的维护策略以维护环境的性能和安全。虽然学术文献中的丰富资源可以支持新型维护解决方案的设计，但鉴于已发表的论文数量巨大，从现有方法中提取可行的见解具有挑战性。此外，现有的评论论文可以帮助总结最先进的技术，将其研究范围限制在特定场景下某些组件的维护。这项工作通过更广泛的文献分析填补了这一空白，其中涵盖了针对云、边缘和物联网环境中的物理和逻辑组件的维护策略。首先，我们引入一种分类法，根据几个特征来组织现有的解决方案。然后，我们按照分类结构回顾文献，以促进对研究前景的理解和现有作品之间的比较。最后，我们揭示了代表有前途的研究方向的开放挑战。