Computer system and availability method thereof

• US 8,874,952 B2
• Filed: 03/05/2010
• Issued: 10/28/2014
• Est. Priority Date: 01/05/2010
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Tameshige, Takashi, Takamoto, Yoshifumi, Hatasaki, Keisuke
• Primary Examiner(s): Ehne, Charles
• Application Number:
  US13/388,948
• Publication Number:
  US 20120131379A1
• Time in Patent Office: 1,698 Days
• US Class Current: 714/4.1
• CPC Class Codes: G06F 9/5077 Logical partitioning of resources; Management or configuration of virtualized resources specific details on emulation or internal functioning of virtual machines G06F9/455
• CPC Class Codes: Y02D 10/00 Energy efficient computing, e.g. low power processors, power management or thermal management

Abstract

High availability computer system and fault correction method. If a fault occurs in the current-system physical device allocated to the current-system virtual device of the virtual server, the virtualization mechanism of the physical server configures, for the standby-system virtual device of the virtual server, the standby-system physical device, as a physical device which is used at a high priority, and the virtualization mechanism distributes the request issued from the standby-system virtual device of another virtual server to a standby-system physical device, but, when such a standby-system physical device does not exist, the virtualization mechanism distributes the request to a standby-system physical device configured for high priority usage.

Claims

1. A computer system comprising at least one physical server,

wherein the at least one physical server comprises:

current-system physical devices;

standby-system physical devices; and

a virtualization mechanism which provides a plurality of virtual servers each comprising a current-system virtual device obtained by virtualizing the current-system physical devices, and a standby-system virtual device obtained by virtualizing the standby-system physical devices,

wherein the virtualization mechanism is configured to;

allocate, for occupied usage, the current-system physical devices to the current-system virtual devices of each of the plurality of virtual servers; and

allocate, for shared usage, the standby-system physical devices to the standby-system virtual devices of the plurality of virtual servers,

wherein, when a fault occurs in a first current-system physical device allocated to a first current-system virtual device of a first virtual server,

the virtualization mechanism is configured to;

configure a first standby-system physical device, which has not been configured as a physical device which is used at high priority by any of the standby-system virtual devices of the plurality of virtual servers, as a physical device which is used at a high priority by a first standby-system virtual device of the first virtual server;

distribute a request issued from the first standby-system virtual device of the first virtual server to the first standby-system physical device configured as a physical device which is used at a high priority by the first standby-system virtual device of the first virtual server; and

distribute a request issued from another standby-system virtual device of another virtual server to another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers when the another standby-system physical device exists.

2. The computer system according to claim 1,

wherein, when the another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers does not exist, the virtualization mechanism is configured to distribute the request issued from the another standby-system virtual device of the another virtual server to a standby-system physical device which is configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers.

3. The computer system according to claim 1, further comprising a management server configured to manage the at least one physical server,

wherein, when the request is issued from the another standby-system virtual device of the another virtual server and the another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers does not exist, the virtualization mechanism is configured to issue an inquiry to the management server to determine a standby-system physical device to distribute the request,

wherein the management server is configured to:

determine the standby-system physical device to distribute the request on a basis of priority rankings of tasks executed by each of the of virtual servers; and

notify the virtualization mechanism of the determination result, and

wherein, on the basis of the determination result of the management server, the virtualization mechanism is configured to distribute the request to a standby-system physical device which is configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers.

4. The computer system according to claim 3,

wherein, when the request is issued from the another standby-system virtual device of the another virtual server, if the another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers does not exist and when there is a standby-system physical device which is capable of being assigned the request and is configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers, the virtualization mechanism is configured to distribute the request to the standby-system physical device which is capable of being assigned the request and,

when the standby-system physical device which is capable of being assigned the request does not exist, the virtualization mechanism is configured to issue the inquiry to the management server to determine a standby-system physical device to be distributed the request.

5. The computer system according to claim 1, further comprising a management server configured to manage the at least one physical server and a storage apparatus for providing a storage area for reading and writing data from the at least one physical server,

wherein the current-system physical devices and standby-system physical devices are adapters for performing communications with the management server or the storage apparatus.

6. The computer system according to claim 1,

wherein the virtualization mechanism is configured to determine whether the fault has occurred in the first current-system physical device allocated to the first current-system virtual device of the first virtual server based on a request type issued by the first standby-system virtual device of the first virtual server.

7. The computer system according to claim 6,

wherein, when the request issued from the first standby-system virtual device of the first virtual server is a vital monitoring request or when a packet size of the request issued from the first standby-system virtual device is less than a threshold, the virtualization mechanism is configured to determine that the fault has occurred in the first current-system physical device allocated to the first current-system virtual device of the first virtual server.

8. The computer system according to claim 7,

wherein, when the request issued from the another standby-system virtual device of the another virtual server is a vital monitoring request and the another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers exists, the virtualization mechanism is configured to distribute the request to the another standby-system physical device, and

wherein, when the another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers does not exist, the virtualization mechanism is configured to distribute the request to a standby-system physical device which is configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers executing a task with a lowest priority ranking, or to a standby-system physical device with a lowest load.

9. An availability method for a computer system comprising at least one physical server,

wherein the at least one physical server comprises:

current-system physical devices;

standby-system physical devices; and

a virtualization mechanism which provides a plurality of virtual servers each comprising a current-system virtual device obtained by virtualizing the current-system physical devices, and a standby-system virtual device obtained by virtualizing the standby-system physical devices,

the availability method comprising;

a first step in which the virtualization mechanism allocates, for occupied usage, the current-system physical devices to the current-system virtual devices of each of the plurality of virtual servers and which allocates, for shared usage, the standby-system physical devices to the standby-system virtual devices of the virtual servers; and

a second step in which, when a fault occurs in a first current-system physical device allocated to a first current-system virtual device of a first virtual server,

the virtualization mechanism configures a first standby-system physical device, which has not been configured as a physical device which is used at high priority by any of the standby-system virtual devices of the plurality of virtual servers, as a physical device which is used at a high priority by a first standby-system virtual device of the first virtual server;

distributes a request issued from the first standby-system virtual device of the first virtual server to the first standby-system physical device configured as a physical device which is used at a high priority by the first standby-system virtual device of the first virtual server; and

distributes a request issued from another standby-system virtual device of another virtual server to another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers when the another standby-system physical device exists.

10. The availability method for a computer system according to claim 9,

wherein, in the second step,

when the another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers does not exist, the virtualization mechanism distributes the request issued from the another standby-system virtual device of the another virtual server to a standby-system physical device which is configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers.

11. The availability method for a computer system according to claim 9,

wherein the computer system further comprises:

a management server configured to manage the physical servers, and

wherein the second step comprises;

an inquiry step in which, when the request is issued from the another standby-system virtual device of the another virtual server and the another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers does not exist, the virtualization mechanism issues an inquiry to the management server to determine a standby-system physical device to distribute the request;

a determination step in which the management server determines the standby-system physical device to distribute the request on a basis of priority rankings of tasks executed by each of the plurality of virtual servers, and notifies the virtualization mechanism of the determination result; and

a distribution step in which, on the basis of the determination result of the management server, the virtualization mechanism distributes the request to a standby-system physical device which is configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers.

12. The availability method for a computer system according to claim 11,

wherein, in the inquiry step,

when the request is issued from the another standby-system virtual device of the another virtual server, if the another standby-system physical device which has not been configured as a physical device which is used at a high priority by any of the standby-system virtual devices of the plurality of virtual servers does not exist and when there is a standby-system physical device which is capable of being assigned the request and is configured as a physical device which is used at a high priority by any of the standby-system virtual device of the plurality of virtual servers, the virtualization mechanism distributes the request to the standby-system physical device which is capable of being assigned the request and,

when the standby-system physical device which is capable of being assigned the request does not exist, the virtualization mechanism issues an inquiry to the management server to determine a standby-system physical device to distribute the request.

13. The availability method of the computer system according to claim 9,

wherein the computer system further comprises

a management server configured to manage the physical server and a storage apparatus for providing a storage area for reading and writing data from the at least one physical server,

wherein the current-system physical devices and standby-system physical devices are adapters for performing communications with the management server or the storage apparatus.

14. The availability method of the computer system according to claim 9,

wherein, in the second step,

the virtualization mechanism determines whether the fault has occurred in the first current-system physical device allocated to the first current-system virtual device of the first virtual server based on a request type issued by the first standby-system virtual device of the first virtual server.

15. The availability method of the computer system according to claim 14,

wherein, in the second step,

the virtualization mechanism determines whether the fault has occurred in the first current-system physical device allocated to the first current-system virtual device of the first virtual server when the request issued from the first standby-system virtual device of the first virtual server is a vital monitoring request or when a packet size of the request issued from the first standby-system virtual device is less than a threshold.

16. The availability method of the computer system according to claim 15,

wherein, in the second step,

when the request issued from the another standby-system virtual device of the another virtual server is a vital monitoring request,

the virtualization mechanism distributes the request to the another standby-system physical device when the another standby-system physical device which has not been configured as a physical device for high priority usage by any of the standby-system virtual devices of the plurality of virtual servers exists, and

when the another standby-system physical device which has not been configured as a physical device for high priority usage by any of the standby-system virtual devices of the plurality of virtual servers does not exist,

the virtualization mechanism distributes the request to a standby-system physical device configured as a physical device for high priority usage by any of the standby-system virtual devices of the plurality of virtual servers which executes a task with a lowest priority ranking, or to a standby-system physical device with a lowest load.