About

Cloud computing era is here. As more and more computing tasks depend on cloud computing, the availability and reliability of cloud services are becoming more and more important. Unfortunately, major reliability and availability problems still take place even recently. Some cloud infrastructures experienced major outages. Some customer data were permanently lost. To sum up, it seems that the computing forecast for tomorrow is "cloudy with a chance of failure."

CloudFail is a joint project between UC Berkeley and UW Madison. We are interested in any issues that relate to cloud computing and failures. Our goal is to investigate fundamental problems of large-scale failure recovery and eventually build more reliable cloud services.

Download

Our PreFail source code is available! Please download it from the download page.

Papers

• Failure as a Service (FaaS): A Cloud Service for Large-Scale, Online Failure Drills [TR-UCB '11]
  Haryadi S. Gunawi, Thanh Do, Joseph M. Hellerstein, Ion Stoica, Dhruba Borthakur, and Jesse Robbins
  
  
• PreFail: A Programmable Tool for Multiple-Failure Injection [OOPSLA '11]
  Pallavi Joshi, Haryadi S. Gunawi, and Koushik Sen
  
  
• PreFail: Programmable and Efficient Failure Testing Framework [FAST '11 WIP and Poster]
  Pallavi Joshi, Haryadi S. Gunawi, and Koushik Sen
  
  
• FATE and DESTINI: A Framework for Cloud Recovery Testing [NSDI '11]
  Haryadi S. Gunawi, Thanh Do, Pallavi Joshi, Peter Alvaro, Joseph M. Hellerstein, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau, Koushik Sen, and Dhruba Borthakur
  
  
• Towards Automatically Checking Thousands of Failures with Micro-specifications [HotDep '10]
  Haryadi S. Gunawi, Thanh Do, Pallavi Joshi, Joseph M. Hellerstein, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau, and Koushik Sen
  
  

People

Sponsors, Collaborators, and Supports

We would like to thank:

• National Science Foundation, Grants# CCF-1016924, CCF-1017073, CCF-1018729, and CCF-0747390
• NSF/CRA/CCC Computing Innovation Fellows Project
• Facebook
• Emulab Network Emulation Testbed