Virtual computer system having SR-IOV compliant device mounted thereon and failure detection method

摘要

A failure detection method including: detecting, by a virtual computer, occurrence of the failure in a virtual function of an I/O device; acquiring, a virtual device name corresponding to the virtual function in which the failure has occurred; referring, to device information retaining a virtual device name of the I/O device assigned to the virtual computer and VF specific information on the I/O device, thereby acquiring the VF specific information based on the acquired virtual device name; transmitting, the acquired VF specific information to the host; referring, by the host, to I/O correspondence information retaining a slot number of a slot in which the I/O device is mounted, and VF specific information, thereby acquiring the slot number corresponding to the VF specific information received from the virtual computer; and identifying, the acquired slot number as the slot number of the I/O device on which the failure has occurred.

主权项

1. A virtual computer system, comprising:
a physical computer including a processor, a memory, at least one slot, and an I/O device mounted in the at least one slot; and a host including a virtualization module for assigning a computer resource of the physical computer to at least one virtual computer, wherein: the I/O device includes a physical function and a virtual function, and is compliant with an SR-IOV; the physical computer uses the physical function of the I/O device via a PF driver of the host for using the physical function, and the at least one virtual computer uses the virtual function of the I/O device via a VF driver for using the virtual function; the at least one virtual computer includes:
device information for retaining a virtual device name of the I/O device assigned to the at least one virtual computer and VF specific information on the I/O device on the memory; anda failure detection module for guest for detecting occurrence of a failure in the virtual function, and notifying the host of the occurrence of the failure; the host includes:
I/O correspondence information for retaining a slot number of a slot in which the I/O device is mounted, and the VF specific information on the memory; anda failure detection module for host for identifying the slot number of the I/O device on which a failure has occurred; when the failure detection module for guest detects the failure of the virtual function, the failure detection module for guest refers to the device information containing the virtual device name and the VF specific information, thereby acquiring the virtual device name of the virtual device providing the virtual function on which the failure has occurred, refers to the device information by the virtual device name, thereby acquiring the VF specific information corresponding to the acquired virtual device name, and transmits the acquired VF specific information to the failure detection module for host; and when the failure detection module for host receives the VF specific information from the failure detection module for guest, the failure detection module for host refers to the I/O correspondence information, thereby identifying the slot number corresponding to the received VF specific information.