Pas-Ca: A Cloud Computing Auto-Scalability Method For High-Demand Web Systems

This work proposes an auto scaling method for high demanding web applications in a cloud computing system. The proposal has the goal to increase efficiency in containers allocation for processing web requests considering a threshold for applications response time. The method was developed using a control algorithm that applies adjustments based on the properties extracted from workload characterization. The method was evaluated with different types of workloads and the results show that the proposal achieves good results by allocating containers more efficiently than other solutions already well-known in the cloud computing market.