Making Cloud Spot Instance Interruption Events Visible
Keywords: cloud computing, spot instance, interruption modeling, enhancing reliability, spot instance datasets
Abstract: Public cloud computing vendors offer a surplus of computing resources at a cheaper price with a service of spot instance. Despite the possible great cost savings from using spot instances, sudden resource interruption can happen as demand changes. To help users estimate cost savings and the possibility of interruption when using spot instances, vendors provide diverse datasets. However, the effectiveness of the datasets is not quantitatively evaluated yet, and many users still rely on hunch when choosing spot instances. To help users lower the chance of interruption of the spot instance for reliable usage, in this paper we thoroughly analyze various datasets of the spot instance and present the feasibility for value prediction. Then, to measure how the public datasets reflect real-world spot instance interruption events, we conduct extensive tests for spot instances of AWS, Azure, and Google Cloud while proposing a novel heuristic to control experimental cost minimizing information loss. Combining the dataset analysis, prediction, and real-world experiment result, we show the feasibility for lowering the possibility of interruption events significantly.
Track: Systems and Infrastructure for Web, Mobile, and WoT
Submission Guidelines Scope: Yes
Submission Guidelines Blind: Yes
Submission Guidelines Format: Yes
Submission Guidelines Limit: Yes
Submission Guidelines Authorship: Yes
Student Author: Yes
Submission Number: 1368
Loading