Burst-Aware Predictive Autoscaling for Containerized Microservices

Autoscaling methods are used for cloud-hosted applications to dynamically scale the allocated resources guaranteeing Quality-of-Service (QoS). The public-facing application serves dynamic workloads, which contain bursts and pose challenges autoscaling ensure performance. Existing State-of-the-art burst-oblivious determine provision appropriate resources. For it is hard detect handle online maintaining In this article, we propose a novel burst-aware method detects burst in workloads using workload forecasting, resource prediction, scaling decision making while minimizing response time service-level objectives (SLO) violations. We evaluated our approach through trace-driven simulation, multiple synthetic realistic bursty containerized microservices, improving performance when comparing against existing state-of-the-art methods. Such experiments show an increase of $\times $×1.09 total processed requests, reduction xlink:href="iqbal-ieq2-2995937.gif"/>5.17 SLO violations, xlink:href="iqbal-ieq3-2995937.gif"/>0.767 cost as compared baseline method.