Maintaining SmartX multi-view visibility for OF@TEIN plus distributed cloud-native edge boxes

Abstract

Maintaining knowledge of infrastructure has always been a key element to ensure the performance of complex distributed systems, being the first step to control quality of service, identify key performance bottlenecks, or make decisions about resource allocation and job scheduling, to name a few. In cloud-native edge boxes, edge computing leverages cloud-native applications to process/analyze some of their computing/storage own close to the location where the data is generated by a large number of heterogeneous devices. Some of the differences between this approach and more conventional architecture, like cloud are that in a distributed environment, these resources/applications have varying running conditions depending on resource availability, quality of network connection, and being geo-distributed. These aforementioned challenges in operations, however, contribute to a requirement toward identifying efficient methods for monitoring operational environment with continuity in order to maintain and sustain services. In this article, we present setup and configuration of distributed resources with cloud-native edge capabilities and having centralized connectivity in a specialized multi-site cloud testbed. Second, we propose a solution to persistently maintain multi-view (ie, resource-layer and flow layer) visibility data (monitoring) collection under varying situations and environments. Third, we present verification results on a prototype implementation and interactive visualization support. This solution allows maintaining monitoring visibility data that conform to the specific demands of cloud-native edge computing resources. We evaluate the research by maintaining visibility data of multi-site cloud at 14 research institutes for several months and we demonstrate that it fulfills the requirements we previously enumerated.