Fault-tolerant computer system, fault-tolerant computer system control method and recording medium storing control program for fault-tolerant computer system

摘要

A fault-tolerant computer system that includes a computer including a virtual machine on which a guest operating system (OS) is operating and a computer in operation, the computer including a snapshot manager that saves a guest OS snapshot, which includes difference information at each checkpoint, of the guest OS in a memory of the computer, and sends an instruction at each checkpoint to copy the guest OS snapshot as a guest OS snapshot to a memory of a computer via a transfer unit, and that computer includes a snapshot manager that activates a guest OS based on the guest OS snapshot when the computer stops.

主权项

1. A fault-tolerant computer system including:
a first computer that comprises a first memory and a first transfer unit, and operates a virtual machine comprising a guest Operating System (OS); and a second computer that comprises a second memory and a second transfer unit that receives data that is transferred from the first transfer unit, wherein the first computer comprises a first snapshot manager that, together with acquiring a snapshot of the virtual machine at each predetermined timing at the first computer, saves, in the first memory, difference information of a snapshot of the virtual machine at a first timing from one previous timing, and sends an instruction to the first transfer unit to transfer the difference information that is saved in the first memory to the second memory via the first transfer unit and the second transfer unit, wherein the second computer comprises a second snapshot manager that, together with generating a snapshot based on the difference information that was transferred to the second memory via the first transfer unit and the second transfer unit and saving the snapshot generated based on the difference information in the second memory, activates the guest OS in the second computer at a predetermined timing at the second computer based on the snapshot that was saved in the second memory, wherein the first computer comprises a first error handler that acquires an error information of the first computer, and sends an instruction to the first transfer unit to transfer the error information to the second computer, the second computer comprises a second error handler that, based on the error information that was sent via the first transfer unit and received via the second transfer unit, determines whether there is a failure in the first computer, and when it is determined that there is a failure, sends a computer switch notification to the second snapshot manager, and the second timing comprises a time when the second snapshot manager receives the computer switch notification, and wherein the error information includes information that indicates there is a possibility that the first computer will stop.