Welcome to the 9Lives Wiki

9Lives is a research project at Sandia National Laboratories to investigate critical changes required in MPP systems software to support system-directed resilience. The goal is to provide efficient, application-transparent resilience through coordinated use of system resources. In particular, this project will research and develop technologies for:

  • Application Quiescence: the ability to suspend CPU, network, and storage services used by an individual application without interfering wiht the progress of other applications;
  • State Management: the ability to identify, extract, and manage application state in a transparent, efficient, and non-intrusive way; and
  • Fault Recovery: the ability to transparently replace a failed component without restarting the entire application.

This project is different than other related efforts (such as BLCR?) because it focuses primarily on the requirements of capability-based platforms, where performance and scalability are critical. Our software will build upon design principles of the lightweight kernel projects at Sandia laboratories, and we will integrate our software into the  Kitten operating system.

9lives Stuff

For a complete list of local wiki pages, see TitleIndex.