Continuous Deployment in IoT Edge Computing : A GitOps implementation

Edge computing (EC) is a recent computing paradigm in the IoT domain for performing calculations and processing at the edge of the network, closer to the user instead of moving all data to the cloud. In this way, EC practitioners aim to replicate the benefits of cloud computing but closer to the source of the data, reducing the latency imposed by the distance between the cloud and the end devices, the network overhead, and the security and privacy risk that this offloading entails. But not only that. EC practitioners also aim to replicate best practices used on cloud-native development and operation. In this regard, GitOps is one of the main trends in the DevOps ecosystem for Continuous Deployment (CD) that promotes infrastructure automation in highly distributed applications, accelerating value delivery while avoiding error-prone manual processes. Till now, GitOps has been mainly related to Kubernetes—a widely used container orchestration platform to run distributed applications resiliently. In this regard, more and more IoT EC solutions are implementing container orchestration solutions like Kubernetes, thus, it is worthwhile identifying the benefits that GitOps adoption could leverage at the Edge and how possible challenges and drawbacks could be addressed. This paper presents a proof of concept based on tools included in the Cloud Native Computing Foundation (CNCF) landscape to check the feasibility of applying GtiOps in IoT EC solutions. Several drawbacks were detected mainly due to the used tooling, such as lack of automatic provisioning of physical infrastructure, tight coupled to Kubernetes descriptors for declaring workloads, and limitations on the edge devices that can be supported.