JP2017045084A - Failure detection apparatus and failure detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To allow a virtual machine to acquire hardware state of a host machine, to properly respond to a failure in the host machine.SOLUTION: A server failure detection function 11 acquires hardware state of host machines 1A, 1B. A failure detection cooperation function 12 transmits the machine state of the host machines 1A, 1B to virtual machines 2A, 2B. A failure detection cooperation function 21 of the virtual machines 2A, 2B receives the machine state of the host machines 1A, 1B, to associate the machine state of the host machines 1A, 1B with that of the virtual machines 2A, 2B. A system running on the virtual machines 2A, 2B acquires the machine state of the host machines 1A, 1B, to detect a failure of the host machines 1A, 1B.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a system failure control technique using a virtual machine operating on a physical machine.

  Carrier grade servers use an HA cluster to maintain fault tolerance, prepare an active server and a standby server, and switch between the standby and active systems when an error occurs in the active server Thus, service continuity is improved (see, for example, Patent Document 1). A system operating on an active server detects the occurrence of an abnormality by acquiring the status of hardware such as a machine on which it is operating and performing status determination.

JP 2014-165632 A

  When a virtual machine is used to build a system that switches between the active system and the standby system by acquiring the status of the hardware on which it is operating and performing status determination, the hardware status acquisition function The status determination function acquires and determines the status of the virtual machine, and even if it continues to detect the status of the virtual machine, it cannot detect hardware-specific temperature abnormalities or failures due to aging, and operate the virtual machine. There was a problem that it was not possible to detect abnormalities in the host machine.

  In general, it is conceivable to deal with this problem by arranging a virtualization layer monitoring device outside to detect the state of the physical machine. However, the external monitoring device cannot detect an abnormality (progressive abnormality or the like) other than the monitoring target function. In addition, when information cannot be transmitted from an external monitoring device to a server program operating on a virtual machine, it is not possible to perform spontaneous machine control of the server program.

  The present invention has been made in view of the above, and an object of the present invention is to make it possible for a virtual machine to acquire a hardware state of a host machine and appropriately cope with a failure of the host machine.

  A failure detection apparatus according to a first aspect of the present invention is a failure detection apparatus in which a virtual machine operates on a physical machine, wherein the physical machine acquires state of the physical machine, the state acquisition means, State communication means for transmitting to the virtual machine, the virtual machine having the state cooperation means for acquiring the state from the physical machine and enabling a program operating on the virtual machine to acquire the state. It is characterized by that.

  The failure detection apparatus is used in a redundant system including an active system and a standby system, and the virtual machine switches between the active system and the standby system when detecting an abnormality of the physical machine. It has a failure control means.

  A failure detection method according to a second aspect of the present invention is a failure detection method by a failure detection device in which a virtual machine operates on a physical machine, the step of acquiring the state of the physical machine by the physical machine, and the state Transmitting the status to the virtual machine, and acquiring the status from the physical machine by the virtual machine and enabling a program operating on the virtual machine to acquire the status. Features.

  According to the present invention, the virtual machine can acquire the state of the hardware of the host machine and can appropriately cope with the occurrence of a failure in the host machine.

It is a functional block diagram which shows the structure of the failure control system in this Embodiment. It is a sequence diagram which shows operation | movement of the failure control system in this Embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a functional block diagram showing the configuration of the fault control system in the present embodiment. The failure control system shown in FIG. 1 is a failure control system in which virtual machines 2A and 2B operating on host machines 1A and 1B execute programs to operate a desired system, and operate on one virtual machine 2A. In this system, the system is the active system, and the system operating on the other virtual machine 2B is the standby system. When an abnormality is detected in the active system, the system is switched to the standby system to continue the service. For example, a telecommunications carrier uses this system for building a communication network. The fault control system has a function of providing a service by operating a program on the virtual machines 2A and 2B. FIG. 1 shows only functions used for fault control.

  The host machines 1A and 1B shown in FIG. 1 have a server failure detection function 11 and a failure detection linkage function 12, and the virtual machines 2A and 2B operating on the host machines 1A and 1B are a failure detection linkage function 21 and a server failure detection. A function 22 is provided.

  The server failure detection function 11 acquires the status of the host machines 1A and 1B using a function of acquiring the machine status of the host machines 1A and 1B. The host machines 1A and 1B are physical machines, and an abnormality occurs or fails in hardware components included in the host machines 1A and 1B. As a function for acquiring the state of the physical machine, for example, the S.D. M.M. A. R. T. T. And memory check.

  The failure detection cooperation function 12 transmits the machine state of the host machines 1A and 1B acquired by the server failure detection function 11 to the virtual machines 2A and 2B, and the system operating on the virtual machines 2A and 2B is the host machine 1A or 1B. It is a function that makes it possible to acquire the machine status.

  As a method of transmitting the machine state, there is a method of mounting an IP layer unique function in the host machines 1A and 1B. Other examples include the following.

(1) When an abnormality occurs, an SNMP trap is used to notify the SNMP manager of the virtual machines 2A and 2B of the abnormality from the SNMP agent of the host machines 1A and 1B.
(2) The heartbeat is transmitted from the virtual machines 2A and 2B to the host machines 1A and 1B, and alive monitoring is performed by reception. If heartbeat does not return, it is detected that there is an abnormality.
(3) The host machines 1A and 1B log in to the virtual machines 2A and 2B, and destroy the virtual machines 2A and 2B at locations where the host machines 1A and 1B are malfunctioning.
(4) A mail server is activated on the host machines 1A and 1B, a mail client is activated on the virtual machines 2A and 2B, and the host machines 1A and 1B transmit to the virtual machines 2A and 2B by mail.
(5) The host machines 1A and 1B make Telnet connections to the virtual machines 2A and 2B, and edit the files on the virtual machines 2A and 2B.
(6) The host machines 1A and 1B are FTP-connected to the virtual machines 2A and 2B, and files are arranged on the virtual machines 2A and 2B.

  On the other hand, the failure detection linkage function 21 of the virtual machines 2A and 2B receives the machine state of the host machines 1A and 1B transmitted by the above method, and links the host machines 1A and 1B with the machine states of the virtual machines 2A and 2B. Thus, the server failure detection function 22 of the virtual machines 2A and 2B can detect the machine state of the host machines 1A and 1B.

  The server failure detection function 22 is a function provided in a system (server program) that operates on the virtual machines 2A and 2B. The server failure detection function 22 determines the status of the virtual machines 2A and 2B and the status of the host machines 1A and 1B. Is detected. Since the server failure detection function 22 performs the state determination of the virtual machines 2A and 2B and the state determination of the host machines 1A and 1B, the operation determination may be performed separately so that the respective states can be understood. Alternatively, the operation determination may be performed by superimposing the state determination of the host machines 1A and 1B on the state determination of the virtual machines 2A and 2B.

  When detecting a failure, the server failure detection function 22 starts a failure control operation for switching to the standby system. Alternatively, an alarm may be output to the maintenance person when an abnormality is detected, and the maintenance person may perform maintenance work.

  Next, the operation of the failure control system in this embodiment will be described.

  FIG. 2 is a sequence diagram showing the operation of the failure control system in the present embodiment.

  The server failure detection function 11 of the host machine 1A acquires the state of the host machine 1A and detects an abnormality (step S11).

  The failure detection cooperation function 12 of the host machine 1A transmits the state of the host machine 1A to the failure detection cooperation function 21 of the virtual machine 2A (step S12).

  If the failure detection cooperation function 21 of the virtual machine 2A determines that it is necessary to switch the active system to the standby system from the state of the host machine 1A, it issues a switching instruction to the virtual machine 2B (step S13).

  As described above, according to the present embodiment, the server failure detection function 11 acquires the hardware state of the host machines 1A and 1B, and the failure detection cooperation function 12 virtualizes the machine state of the host machines 1A and 1B. Are transmitted to the machines 2A and 2B, and the fault detection cooperation function 21 of the virtual machines 2A and 2B receives the machine statuses of the host machines 1A and 1B, and links the host machines 1A and 1B with the machine statuses of the virtual machines 2A and 2B. As a result, the system operating on the virtual machines 2A and 2B can acquire the machine state of the host machines 1A and 1B, and can detect an abnormality in the host machines 1A and 1B.

1A, 1B ... Host machine 11 ... Server failure detection function 12 ... Failure detection cooperation function 2A, 2B ... Virtual machine 21 ... Failure detection cooperation function 22 ... Server failure detection function

Claims (3)

A failure detection device in which a virtual machine operates on a physical machine,
The physical machine is
Status acquisition means for acquiring the status of the physical machine;
State transmitting means for transmitting the state to the virtual machine;
The virtual machine is
A failure detection apparatus comprising: a state cooperation unit that acquires the state from the physical machine and enables a program operating on the virtual machine to acquire the state. The failure detection device is used in a redundant system including an active system and a standby system,
The virtual machine is
The failure detection apparatus according to claim 1, further comprising a failure control unit that switches between an active system and a standby system when an abnormality of the physical machine is detected. A failure detection method by a failure detection apparatus in which a virtual machine operates on a physical machine,
By the physical machine,
Obtaining the state of the physical machine;
Communicating the state to the virtual machine,
By the virtual machine,
A failure detection method comprising: acquiring the state from the physical machine, and enabling a program operating on the virtual machine to acquire the state.

Images