Incendio: Priority-Based Scheduling for Alleviating Cold Start in Serverless Computing

Abstract

Incendio reduces end-to-end serverless response latency through a function-priority model, Prophet-LightGBM-based container prewarming and reclamation, and reinforcement-learning-assisted scheduling. Implemented on Apache OpenWhisk, it improves performance by 1.4× over the native system and reduces latency by 23% and 14.8% compared with two state-of-the-art approaches.

Qianlong Sang

Fifth-Year Computer Science
Ph.D. Student

My research focuses on operating systems, mobile and edge systems, and efficient on-device AI inference.