Abstract: A multi-processor switch includes a main processor, a plurality of co-processors to perform a calling process under the control of the main processor, a detection unit to detect a failure at the main processor, and a determination unit to determine which replacement processor will perform the main processor functions from among the co-processors, based on the operating state of each co-processor when the failure at the main processor is detected by the detection unit. The operating state of each of the co-processor is, for example, the load state imposed on the co-processor or the number of times its operation has been resumed. By selecting a co-processor that has a small load or which has been resumed less frequently, an appropriate, more stable replacement processor can be determined.

Abstract: An apparatus and method utilized to automatically test a computer is provided. By using the data storage medium for testing in which video data and audio data for testing are stored in different positions, output video data and output audio data are produced by applying the video and audio data for testing to a video board and an audio board, respectively, and determining whether the video and audio data for testing and the output video and audio data are the same, thereby testing automatically whether the video board and the audio board are functioning normally.

Abstract: In a state driven machine logically sub-divided into a plurality of blocks, with each block including a sub-set of a total number of process defining operating states and transition paths, plural records each relating to a certain execution are associated with each block. A log is maintained on a per execution basis of events (such as state transitions and actions) which occur for that execution. In one implementation, the log is maintained in each record to store events occurring for the certain execution within the block associated with that record. In another implementation, the log is maintained to store events occurring for the certain execution across all implicated blocks. Using the information stored in the log, it is possible to trace back along process execution to identify an origin of a fault in the process implemented by the state driven machine.