Thesis
Scheduling functions-as-a-service at the edge
Washington State University
Master of Science (MS), Washington State University
2018
Handle:
https://hdl.handle.net/2376/103047
Abstract
Following the recent adoption of FaaS technologies and the introduction of numerous self-hosted FaaS systems, the need for real time monitoring and scheduling of functions in an ecosystem of providers has emerged. In this thesis, we present a novel performance monitoring and multi-provider scheduling framework, Painless Scheduling, for scheduling across FaaS providers. Painless Scheduling monitors the performance of serverless providers in real time such as AWS Lambda and Openwhisk, as well as edge resources. Additionally, Painless Scheduling has the ability to dynamically advise clients on the optimal resource to use based on their unique scheduling needs as conditions change for any reason. In our evaluation of Painless Scheduling framework utilizing two cloud providers and a local provider, Painless Scheduling selected a "near-optimal" schedule. When compared to scheduling on a single cloud resource Painless Scheduling provides a 4X speedup across multiple providers in a volatile edge computing environment. We conclude with future research directions for Painless Scheduling framework and more generally future research in FaaS technology.
Metrics
27 File views/ downloads
92 Record Views
Details
- Title
- Scheduling functions-as-a-service at the edge
- Creators
- Austin M. Aske
- Contributors
- Xinghui Zhao (Degree Supervisor)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; [Pullman, Washington] :
- Identifiers
- 99900525059801842
- Language
- English
- Resource Type
- Thesis