, Data General
Abstract:
Two on-going areas of research:
Predictive Failure Analysis
With the objective of lowering the cost of ownership, research into several different areas. These include both hardware and software predictive failure analysis. Identifying and monitoring key parameters that would predict either failure or performance degradation.
Possible application of a rules based system to analyze failure indicators and suggest an appropriate response to either the customer or an on-line service center. For software, the appropriate response may include automatic on-line repair. For hardware, appropriate response could include throttling down the system to degrade performance but ensure system availability.
Clustering
Application of I2O technology and VIA standards to with Microsoft Clustering Technology (Wolfpack Phase I and Phase II) to obtain enhanced system expandability as well as increased system performance. Investigation into the requirements for an interconnect technology for I2O which would also serve as highly reliable interconnect for Microsoft Clustering as well as other distributed software. Integration of other technology standards such as Hot Plug with I2O and clustering to further enhance system availability.