Method and system to provide high availability of shared data

• US 8,266,472 B2
• Filed: 07/28/2006
• Issued: 09/11/2012
• Est. Priority Date: 05/03/2006
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Bose, Patrick Glen
• Primary Examiner(s): Wilson, Yolanda L
• Application Number:
  US11/460,949
• Publication Number:
  US 20070294563A1
• Time in Patent Office: 2,237 Days
• Field of Search: 714/4, 714/4.11, 714/6.21
• US Class Current: 714/4.11
• CPC Class Codes: G06F 11/2092 Techniques of failing over between control units

Abstract

A data storage system and method are described. The system may comprise a data repository to store data, and first and second storage controllers. The first and second storage controllers may be connected via at least one network device to the data repository. One or more clients may be connected to the first storage controller and to the second storage controller to access the data in the data repository via the first and second storage controllers. The second storage controller may be configured to at least partially inhibit network connectivity of the first storage controller if the first storage controller at least partially fails. The system may include a Storage Area Network providing the data repository, each of the first and second storage controllers being connected to the Storage Area Network via switch. Each switch comprises at least one fiber channel gateway and at least two Target Channel Adaptors.

Claims

1. A data storage system comprising:

a data repository to store data;

an active storage controller coupled to the data repository via at least one network device;

a standby storage controller coupled to the active storage controller and the data repository via the at least one network device, the standby storage controller configured to monitor an active status of the active storage controller to detect failover of the active storage controller, the monitoring to check whether the active storage controller is running and answering requests, the monitoring neither requiring an acknowledgement from the active storage controller nor a status message to be generated by the active storage controller; and

at least one communications link to couple at least one client to the active storage controller and to the standby storage controller to access the data in the data repository via the active storage controller and the standby storage controller, the standby storage controller being configured to become active and at least partially inhibit network connectivity of the active storage controller during the failover, without requiring a host computer to either control the failover or provide switching operations during the failover.

2. The system of claim 1, wherein the active storage controller is a server actively providing the at least one client access to the data repository.

3. The system of claim 1, wherein the active storage controller and the standby storage controller are provided with a common virtual IP address to provide network connectivity to the at least one client.

4. The system of claim 1, further comprising a Storage Area Network providing the data repository, the active storage controller and the standby storage controller each being coupled to the Storage Area Network via at least one switch.

5. The system of claim 4, wherein the at least one switch comprises:

at least one fibre channel gateway to provide network connectivity between the at least one switch and the Storage Area Network; and

at least two Target Channel Adaptors to provide network connectivity to a Host Channel Adaptor of each of the active storage controller and the standby storage controller.

6. The system of claim 5, wherein an initiator is created on the at least one fibre channel gateway to associate a Globally Unique Identifier of the Host Channel Adaptor and a fibre channel addresses of each switch.

7. The system of claim 6, wherein the fibre channel addresses comprise a World Wide Node Name (WWNN) or a World Wide Port Name (WWPN).

8. The system of claim 5, wherein to at least partially inhibit network connectivity of the active storage controller in the event of the failover comprises modifying a zoning configuration on the switch associated with the active storage controller.

9. The system of claim 8, wherein modifying the zoning configuration comprises removing an address of the active storage controller from a defined zone.

10. The system of claim 1, wherein the active storage controller and the standby storage controller each comprises a failover module to monitor operation of the other storage controller.

11. The system of claim 1, which includes a Network File System server connected to the data repository, the active storage controller and the standby storage controller each being connected to the Network File System server via an Ethernet switch.

12. The system of claim 11, wherein the switches are configured in a Virtual Local Area Network via which the active storage controller and the standby storage controller access the Network File System server.

13. The system as claimed in claim 12, wherein to at least partially inhibit network connectivity of the active storage controller in the event of the failover comprises removing a port of the Ethernet switch to which the active storage controller is connected.

14. A method of managing a data repository to store data accessible by at least one client machine, the method comprising:

providing the at least one client with access to the data repository through an active storage controller, the active storage controller being communicatively coupled via at least one network device to the data repository;

monitoring, by a standby storage controller, an active status of the active storage controller to identify at least a partial failure of the active storage controller, the monitoring to check whether the active storage controller is running and answering requests, the monitoring neither requiring an acknowledgement from the active storage controller to the standby storage controller nor a status message to be generated by the active storage controller; and

based on a determination that the active storage controller at least partially failed, providing a failover, the failover including;

modifying network connectivity of the active storage controller utilizing the standby storage controller; and

configuring network connectivity of the standby storage controller to provide the at least one client with access to the data repository without requiring a host computer to either control the failover or provide switching operations during the failover.

15. The method of claim 14, further comprising providing the active and standby storage controllers with a common virtual Internet Protocol (IP) address to provide the network connectivity to the at least one client.

16. The method of claim 14, further comprising associating a Globally Unique Identifier of a Host Channel Adaptor of a server of the data repository with one or more fibre channel addresses of the network device.

17. The method of claim 14, wherein to modify the network connectivity of the standby storage controller comprises removing a port the of an Ethernet switch to which the active storage controller is connected.

18. The method of claim 14, wherein to modify network connectivity of the active storage controller comprises modifying a zoning configuration on a switch associated with the active storage controller.

19. The method of claim 18, wherein modifying the zoning configuration comprises removing an address of the active storage controller from a defined zone.

20. A non-transitory machine-readable storage medium containing instructions that, when executed by one or more processors, perform a method of managing a data repository to store data accessible by at least one client machine, the method comprising:

providing the at least one client machine with access to the data repository through an active storage controller, the active storage controller being communicatively coupled via at least one network device to the data repository;

monitoring, by a standby storage controller, an active status of the active storage controller to identify at least a partial failure of the active storage controller, the monitoring to check whether the active storage controller is running and answering requests, the monitoring neither requiring an acknowledgement from the active storage controller to the standby storage controller nor a status message to be generated by the active storage controller; and

based on a determination that the active storage controller at least partially failed, providing a failover, the failover including;

modifying network connectivity of the active storage controller utilizing the standby storage controller; and

configuring network connectivity of the standby storage controller to provide the at least one client with access to the data repository without requiring a host computer to either control the failover or provide switching operations during the failover.