FAULT TOLERANT SYSTEM, FAULT TOLERANT PROCESSING METHOD AND RECORDING MEDIUM FOR FAULT TOLERANT CONTROL PROGRAM

摘要

PROBLEM TO BE SOLVED: To easily designate a specific CPU that takes over the processing of another faulty CPU for every task in a multiprocessor system. SOLUTION: A take-over decision part 11 searches for a task which is operating with a faulty CPU from the transition information, i.e., the series of the CPU number that takes over the task included in a reconfiguration table 2 and the current transition position showing the CPU under processing in a current transition position table 4 after the CPU fault is detected and then checks the CPU that takes over the task. When the checked CPU is its own one, a standby task start part 12 starts a standby task 3 of the relevant task to take over the processing. A current transition position updating part 13 updates the current transition position of the table 4. When no CPU takes over the task, the task or a total system is discontinued according to the processing selection information.