Strong Partitioning Protocol for VME-based Systems Integration

摘要

This paper describes a new technology that can provide a global VME memory management in supporting the need of strong partitioning among multi- processor applications on the VME backplane. The technology allows each VME board to maintain a fault containment region on board level and prohibit fault propagation through the bus. The described techniques do not require any modification in the standard and the existing boards, and consequently, maintains the plug-and-play advantage of the VMEbus hardware products. The approach is to use a message passing mechanism for multiprocessing instead of shared memory. A software layer has been developed to support message-based inter-module communications using the available features provided by any standard VME card and still detect errors and prevent fault propagation.