Adaptive Micro-service based Cloud Monitoring and Resource Orchestration
Tariq Daradkeh, Anjali Agarwal · 2022
The ultimate goal of cloud management is to get a reliable, mature and automatic elastic model, which can handle workload changes and system resources capacity scaling instantly without SLA violation and with cheapest cost. We propose a new monitoring and orchestration model that is based on micro-services architecture style and presents elasticity scaling measured by four aspects: scaling accuracy, resources orchestration time, monitoring scalability and communication complexity. We achieved an excellent elastic scaling with minimum delay using dynamic 2-3-4 tree structure using integrated self-monitoring and finite state machine configuration. We implemented this model using Java programming language by creating a time events simulator for cloud workload and resources. We validated the model by running real cloud traces and compare our model in monitoring and orchestration of the resources.