Current Internet-of-Things (IoT) systems are highly distributed systems, which integrate cloud, fog, and edge computing approaches. Accelerating their maintenance and continuous improvement, while ensuring their availability, is complex. DevOps promotes fast and continuous feedback from operations to development to detect problems before customers are impacted, among other benefits. However, there is not any formal definition of how to do this. This article defines the “fast and continuous monitoring feedback of system availability” activity (F&CF availability) that supports automatic and continuous monitoring feedback from operations to the development of the IoT system availability. This activity has been formalized through the software and systems process engineering metamodel (SPEM). Its implementation is demonstrated in a real scenario that provides evidence that the formalization of the F&CF availability activity helps teams in better diagnosing and fixing outage problems. The result is a distributed and configurable monitoring component developed through code [monitoring as code (MaC)]. This component is embedded in the IoT infrastructure. MaC enables DevOps team to configure their own metrics and indicators at runtime, i.e., monitoring on demand. The formalization of this activity, based on an MaC technique, enables the automation, versioning, and replication of monitoring elements.

中文翻译：

---

物联网系统的 DevOps：快速、持续地监控系统可用性的反馈


当前的物联网（IoT）系统是高度分布式的系统，集成了云、雾和边缘计算方法。加速其维护和持续改进，同时确保其可用性，是很复杂的。 DevOps 促进从运营到开发的快速、持续的反馈，以便在客户受到影响之前发现问题，以及其他好处。但是，对于如何执行此操作没有任何正式定义。本文定义了“系统可用性的快速持续监控反馈”活动（F&CF 可用性），支持从运营到物联网系统可用性开发的自动持续监控反馈。该活动已通过软件和系统过程工程元模型 (SPEM) 正式化。它的实施在真实场景中进行了演示，证明 F&CF 可用性活动的形式化有助于团队更好地诊断和修复中断问题。结果是通过代码开发的分布式且可配置的监控组件[监控即代码（MaC）]。该组件嵌入在物联网基础设施中。 Mac使DevOps团队能够在运行时配置自己的指标和指标，即按需监控。该活动的形式化基于 Mac 技术，可实现监控元素的自动化、版本控制和复制。