RISE: A Grid-Based Self-Configuring and Self-Healing Remote System Image Management Environment

Autonomic remote system image management support (e.g., remote system installation and backup) is crucial for the management of clusters in large scale grid computing environments in which clusters at different locations are integrated into a single large sale computing system. We propose a Grid-based remote image management environment (called RISE) which is designed to support ease of use, high performance installation, and resiliency against failure and configuration changes. Major features of the RISE system are: (1) remote system image installation, backup and recovery via a web-based simple and intuitive user interface, (2) parallel installation, (3) backup image databases (2) self-configuration (autonomic configuration propagation and recovery protocol), (3) self-healing (autonomic backup image database replication and recovery). The development of the system with no support for grid environments has been completed and its commercial version by Samsung Electronics has been available since the June of 2006. The development of the grid support version is under-way.