System Support for Transient Cloud Computing
In many emerging scenarios, clouds and data centers can only provide transient server availability. This project was the first to develop system support for transience that treated it as a first-class design principle. Since traditional fault-tolerance techniques are too expensive for handling transient server availability, the project developed new transience-aware mechanisms that balanced cost and performance. This project was funded by the National Science Foundation under grant CNS-1422245, a 2016 Google Faculty Research Award, and Amazon (in the form of cloud credits).
Publications
Energy-Agile Design for Parallel HPC Applications
Cong Wang, Michael Zink, and David IrwinSUSCOM 2018HotSpot: Automated Server Hopping in Cloud Spot Markets
Supreeth Shastri and David IrwinSoCC 2017Towards Index-based Global Trading in Cloud Markets
Supreeth Shastri and David IrwinHotCloud 2017The Financialization of Cloud Computing: Opportunities and Challenges
David Irwin, Prateek Sharma, Supreeth Shastri, and Prashant ShenoyICCCN 2017Portfolio-driven Resource Management for Transient Cloud Servers
Prateek Sharma, David Irwin, and Prashant ShenoySIGMETRICS 2017Managing Risk in a Derivative IaaS Cloud
Prateek Sharma, Stephen Lee, Tian Guo, David Irwin, and Prashant ShenoyTPDS 2017Design and Operation Analysis of a Green Data Center
Prateek Sharma, Patrick Pegus, David Irwin, and Prashant ShenoyInternet Computing 2017Keep it Simple: Bidding for Servers in Today’s Cloud Platforms
Prateek Sharma, David Irwin, and Prashant ShenoyInternet Computing 2017Transient Guarantees: Maximizing the Value of Idle Cloud Capacity
Supreeth Shastri, Amr Rizk, and David IrwinSC 2016Cloud Spot Markets are Not Sustainable: The Case for Transient Guarantees
Supreeth Subramanya, Amr Rizk, and David IrwinHotCloud 2016Prateek Sharma, David Irwin, and Prashant ShenoyHotCloud 2016SpotLight: An Information Service for the Cloud
Xue Ouyang, David Irwin, and Prashant ShenoyICDCS 2016Flint: Batch-Interactive Data-Intensive Processing on Transient Servers
Prateek Sharma, Tian Guo, Xin He, David Irwin, Prashant ShenoyEuroSys 2016Analyzing the Efficiency of a Green University Data Center
Patrick Pegus II, Benoy Varghese, Tian Guo, David Irwin, Prashant Shenoy, Anirban Mahanti, James Culbert, John Goodhue, and Chris HillICPE 2016Managing Server Clusters on Intermittent Power
Navin Sharma, Dilip Krishnappa, Sean Barker, David Irwin, and Prashant ShenoyPeerJ 2015Optimizing Parallel HPC Applications for Green Energy Sources
Cong Wang, Michael Zink, and David IrwinIGSC 2015SpotOn: A Batch Computing Service for the Spot Market
Supreeth Subramanya, Tian Guo, Prateek Sharma, David Irwin, and Prashant ShenoySoCC 2015Cutting the Cost of Hosting Online Internet Services using Cloud Spot Markets
Xin He, Ramesh Sitaraman, Prashant Shenoy, and David IrwinHPDC 2015Energy-Agility: A New Grid-centric Metric for Evaluating System Performance
Supreeth Subramanya, Zain Mustafa, David Irwin, and Prashant ShenoyLIMITS 2015SpotCheck: Designing a Derivative IaaS Cloud on the Spot Market
Prateek Sharma, Stephen Lee, Tian Guo, David Irwin, and Prashant ShenoyEuroSys 2015Here Today, Gone Tomorrow: Exploiting Transient Servers in Data Centers
Rahul Singh, Prateek Sharma, David Irwin, Prashant Shenoy, and K.K. RamakrishnanInternet Computing 2014BlinkFS: A Distributed File System for Intermittent Power
Navin Sharma, David Irwin, and Prashant ShenoySUSCOM 2014A Distributed File System for Intermittent Power
Navin Sharma, David Irwin, and Prashant ShenoyIGCC 2013Yank: Enabling Green Data Centers to Pull the Plug
Rahul Singh, David Irwin, Prashant Shenoy, and K.K. RamakrishnanNSDI 2013