Robert Bosch

Harald Buhren

Stefan Koch

Karl-Heinrich Preis

Bernd Klumpp

Helmut Geselle

A method is proposed for monitoring a computer network having at least two subscribers connected via a databus comprising at least two lines, said subscribers having in each case one receiving component and/or one transmitting component. The method is characterized by the fact that the functioning of the databus and/or the functioning of the subscriber is monitored with the aid of error detection signals (watch-dog signals, CEI) of at least one subscriber, and emergency operation measures (M1; M2), matched to the respective error case, for setting defined emergency modes of operation (single-wire mode, special emergency mode) are taken. After taking the first emergency operation measure (M1), it is tested whether the computer network is functioning free of errors. If this is the case, the first emergency operation measure is maintained and thus also the emergency mode of operation (single-wire emergency mode) adopted. If further errors should arise in the computer network after the first emergency operation measure has been taken, the first emergency operation measure is cancelled and the second emergency operation measure (M2) is taken. If, from now on, no further errors arise, the emergency operation mode, the special emergency mode, set by this measure is maintained. Only if further errors continue to arise are the affected subscribers or the entire computer network switched off (bus-off).

Monitoring a computer network

The invention is directed to a multi-processor system which includes at least two processors. The system carries out a first sequence of steps when power is switched on and, for a restart during operation, executes a second sequence of steps. The system distinguishes between power on and a restart and selects, the corresponding step sequence. In addition, a check is provided as to whether the number of resets of a processor has exceeded a pregiven threshold. In this case, a processor is transferred into the standby state for the operating cycle which is then running.

Multi-processor system in a motor vehicle

Process and device for monitoring a set value indicator of a drive motor

The invention is directed to an arrangement for monitoring a computer system with two processors in a motor vehicle. The arrangement provides that the two processors monitor each other to the same extent and that this monitoring occurs in the form of a hand-shake operation in the context of a cyclical data exchange between both processors. In this connection, it is especially advantageous that both processors: can operate fully independently of each other; need not utilize a directly coupled or common I/O-bus; and, are mutually asynchronously operable with different clock frequencies. The arrangement provides a mutual monitoring of two processors to the same extent so that a failure can be localized when it occurs. In this way, the arrangement can distinguish whether a processor has failed or whether an error is present in the peripheral hardware of the processors. For this purpose, the above-mentioned cyclical data exchange between both processors is utilized as an acknowledgement path for the monitoring function.

Growing community of inventors

Buhl, Germany

Harald Buhren