Snape: Reliable and Low-Cost Computing with Mixture of Spot and On-Demand VMs
Fangkai Yanga, Lu Wanga, Zhenyu Xua, Jue Zhanga, Liqun Lia, Bo Qiaoa, Camille Couturierb, Chetan Bansalc, Soumya Ramd, Si Qina, Zhen Mac, Íñigo Goirid, Eli Cortezd, Terry Yange, Victor Rühlef, Saravan Rajmohanc, Qingwei Lina, Dongmei Zhanga
- aMicrosoft Research, China
- bMicrosoft 365, France
- cMicrosoft 365, USA
- dMicrosoft Azure, USA
- eMicrosoft 365, China
- fMicrosoft 365, UK
Papers open inside SciRah from the original publisher or repository; SciRah never hosts them.
Abstract
Cloud providers often have resources that are not being fully utilized, and they may offer them at a lower cost to make up for the reduced availability of these resources. However, customers may be hesitant to use such offerings (such as spot VMs) as making trade-offs between cost and resource availability is not always straightforward. In this work, we propose Snape (Spot On-demand Perfect Mixture), an intelligent framework to optimize the cost and resource availability by dynamically mixing on-demand VMs with spot VMs. Through a detailed characterization based on real production traces, we verify that the eviction of spot VMs is predictable to some extent. Snape also leverages constrained reinforcement learning to adjust the mixture policy online. Experiments across different configurations show that Snape achieves 44% savings compared to using only on-demand VMs while maintaining 99.96% availability, which is 2.77% higher than using only spot VMs.