Service Demand Distribution Estimation for Microservices Using Markovian Arrival Processes.

Building performance models for microservices applications in DevOps is costly and error-prone. Accurate service demand distribution estimation is critical to performance model parameterization. However, traditional service demand estimation methods focus on capturing the mean service demand, disregarding higher-order moments of the distribution. To address this limitation, we propose to estimate higher moments of the service demand distribution for a microservice from monitoring traces. We first generate a closed queueing model to abstract a microservice and model the departure process at the queue node as a Markovian arrival process. This allows formulating the estimation of service demand as an optimization problem, which aims to find the optimal parameters of the first multiple moments of the service demand distribution based on the inter-departure times. We then estimate the service demand distribution with a novel maximum likelihood algorithm, and heuristics to mitigate the computational cost of the optimization process for scalability. We apply our method to real traces from a microservice-based application and demonstrate that its estimations lead to greater prediction accuracy than exponential distributions assumed in traditional service demand estimation approaches.