Cloud cover : monitoring large-scale clouds with Varanus

This research was supported by a Royal Society Industry Fellowship and an Amazon Web Services (AWS) grant.Cloud computing has quickly become the de facto means to deploy large scale systems in a robust and cost effective manner. Central to the maintenance of large scale systems is monitoring which allows for the detection of faults, errors and anomalies and the enacting of optimisation and corrective measures. Monitoring large scale systems is significant challenge requiring the low latency movement of large volumes of data and near real time analysis. This challenge is magnified by elasticity and other cloud properties which previous monitoring systems do not yet fully account for. In this paper we propose Varanus 1 a cloud aware monitoring tool that provides robust, fault tolerant monitoring at scale. We describe in detail the mechanisms which enable Varanus to function effectively and explore the performance of Varanus through a detailed evaluation.Publisher PDFPeer reviewe