Data mirroring using a virtual connection

• US 8,903,766 B2
• Filed: 10/31/2005
• Issued: 12/02/2014
• Est. Priority Date: 10/31/2005
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Padovano, Michael, Christian, Charles E. Jr., Clemens, Laura, Daniels, Rodger, Levinson, Deborah, Reuter, James Michael, Spence, Susan, Stroberger, Christopher, Byrnes, Michael G., Fraser, Donald R.
• Primary Examiner(s): Bhatia, Ajay
• Assistant Examiner(s): MINA, FATIMA P
• Application Number:
  US11/264,571
• Publication Number:
  US 20070100909A1
• Time in Patent Office: 3,319 Days
• Field of Search: 707/204, 707/633
• US Class Current: 707/610
• CPC Class Codes: G06F 11/2058 using more than 2 mirrored copies
• CPC Class Codes: G06F 11/2069 Management of state, configuration or failover
• CPC Class Codes: G06F 11/2071 using a plurality of controllers
• CPC Class Codes: G06F 11/2082 Data synchronisation

Abstract

In a storage system, at least three storage sites are arranged in a full mirroring relationship using a virtual connection between instances of a storage group on two of the sites whereby no user data is transmitted.

Claims

1. A data store apparatus comprising:

a controller that manages data storage among at least three sites, each site comprising a hard drive, arranged in a mirroring configuration, the controller maintaining a full mirroring relationship using a virtual connection between instances of a storage group on two of the at least three sites whereby no user data is transmitted on the virtual connection and data on the virtual connection is monitored through a third site.

2. The data store apparatus according to claim 1 further comprising:

the controller managing data storage in a cascaded mirroring configuration that sends data synchronously from a site A to a site B and then asynchronously from the site B to a site C, the controller maintaining the virtual connection from the site A to the site C and monitoring data that the site B sends to the site C, enabling the site A to establish a mirror relationship with the site C in event of site B failure.

3. The data store apparatus according to claim 1 further comprising:

the controller managing data storage in a fan-out mirroring configuration that sends data from a site A to a site B and a site C, the controller maintaining the virtual connection that enables the site B to monitor data that the site A sends to the site C, enabling the site B to establish a mirror relationship with the site C in event of site A failure.

4. The data store apparatus according to claim 1 further comprising:

the controller creating the virtual connection upon receipt of an administrative command specifying virtual connection attributes including source site designation, destination site designation, and information site designation, the information site informing the source site of data transmission to the destination site.

5. The data store apparatus according to claim 1 further comprising:

the controller creating the virtual connection upon receipt of an administrative command specifying virtual connection attributes including designation of a source instance, designation of a destination instance, and designation of an information instance that informs the source instance of data transfer to the destination instance, the controller creating the virtual connection between the source instance and the destination instance.

6. The data store apparatus according to claim 5 further comprising:

the controller causing the source instance to send an initial message to the information instance, the initial message physically identifying sites of the source, destination, and information instances; and

the controller causing the information instance to verify position in the mirror configuration and to respond to the source instance indicating success for a successful configuration and indicating failure for unsuccessful configuration.

7. The data store apparatus according to claim 6 further comprising:

the controller causing the source instance to respond to successful configuration of the information instance by queuing entries for the destination instance as a target of an asynchronous mirroring relationship; and

the controller causing the source instance to receive a request, write to local virtual disks according to the request, create a data packet containing at least data to be mirrored, uniquely identify the data packet, send the data packet to the information instance, wait for information instance acknowledgement of successful writing to information instance virtual disks, and queue the data packet local to the source instance.

8. The data store apparatus according to claim 7 further comprising:

the controller causing the information instance to receive the data packet from the source instance, write the data packet to the information instance virtual disks, return acknowledgement of data packet successful writing to the source instance, and queue data packets to be sent to the destination instance local to the information instance; and

the controller causing the information instance to send the queued data packets to the destination instance, wait for destination instance acknowledgement of successful writing to destination instance virtual disks, remove data packets from the queue in response to destination instance acknowledgement of writing the data packets, and return acknowledgement of data packet successful writing in the destination instance to the source instance.

9. The data store apparatus according to claim 8 further comprising:

the controller causing the destination instance to receive the data packet from the information instance, write the data packet to the destination instance virtual disks, record identification of the written data packet, and return acknowledgement of data packet successful writing to the information instance.

10. The data store apparatus according to claim 7 further comprising:

the controller causing the information instance to respond to a full queue comprising removing individual data elements from the queue, setting a bit in a bitmap corresponding to an individual data element, setting a bit in the bitmap corresponding to new input/output (I/O) requests, associating with the bitmap a unique identifier (ID) of a last data element placed in the bitmap, placing new data elements in the queue, removing all data elements from the queue and setting a bit in the bitmap corresponding to the removed data elements on detection of a full queue condition, and continuing queuing the data packets local to the source instance.

11. A data system comprising:

at least three storage sites, each site comprising a hard drive, arranged in a full mirroring relationship using a virtual connection between instances of a storage group on two of the sites whereby no user data is transmitted on the virtual connection and data on the virtual connection is monitored through a third site.

12. The data system according to claim 11 further comprising:

a management model enabling a user to replicate a consistency group containing a set of zero or more virtual disks.

13. The data system according to claim 11 further comprising:

a logic executable at the storage sites that creates a virtual connection upon receipt of an administrative command that specifies virtual connection attributes including a source instance designation, a destination instance designation, and an information instance designation, the virtual connection adapted to ensure data transfer from a source instance to a destination instance.

14. The data system according to claim 13 further comprising:

a logic executable at the storage sites that causes the source instance to send an initial message to the information instance, the initial message physically identifying storage sites associated with the source, destination, and information instances;

the logic causing the information instance to verify position in the mirror configuration and to respond to the source instance indicating success for a successful configuration and indicating failure for unsuccessful configuration.

15. The data system according to claim 14 further comprising:

a logic executable at the storage sites that causes the source instance to respond to successful configuration of the information instance by queuing entries for the destination instance as a target of an asynchronous mirroring relationship; and

the logic causing the source instance to receive a request, write to local virtual disks according to the request, create a data packet containing at least data to be mirrored, uniquely identify the data packet, send the data packet to the information instance, wait for information instance acknowledgement of successful writing to information instance virtual disks, and queue the data packet local to the source instance.

16. The data system according to claim 15 further comprising:

a logic executable at the storage sites that causes the information instance to receive the data packet from the source instance, write the data packet to the information instance virtual disks, return acknowledgement of data packet successful writing to the source instance, and queue data packets to be sent to the destination instance local to the information instance; and

the logic causing the information instance to send the queued data packets to the destination instance, wait for destination instance acknowledgement of successful writing to destination instance virtual disks, remove data packets from the queue in response to destination instance acknowledgement of writing the data packets, and return acknowledgement of data packet successful writing in the destination instance to the source instance.

17. The data system according to claim 16 further comprising:

a logic executable at the three storage sites that causes the destination instance to receive the data packet from the information instance, write the data packet to the destination instance virtual disks, record identification of the written data packet, and return acknowledgement of data packet successful writing to the information instance.

18. A method for handling data in a storage system comprising:

arranging at least three sites, each site comprising a hard drive, in a mirroring configuration;

managing data storage among the at least three sites in a full mirroring relationship; and

creating a virtual connection between instances of a storage group on two of the sites whereby no user data is transmitted on the virtual connection and data on the virtual connection is monitored through a third site.

19. The method according to claim 18 further comprising:

managing data storage in a cascaded mirroring configuration;

sending data synchronously from a site A to a site B and then asynchronously from the site B to a site C;

maintaining the virtual connection from the site A to the site C; and

monitoring data that the site B sends to the site C whereby the site A is enabled to establish a mirror relationship with the site C in event of site B failure.

20. The method according to claim 18 further comprising:

managing data storage in a fan-out mirroring configuration that sends data from a site A to a site B and a site C;

maintaining the virtual connection at the site B;

monitoring at the site B data that the site A sends to the site C; and

establishing at the site B a mirror relationship with the site C in event of site A failure.

21. A non-transitory, computer readable medium, comprising

computable readable program code embodied therein adapted to handle data in a storage system, the computable readable program code further comprising:

a code adapted to cause the controller to arrange at least three sites, each site comprising a hard drive, in a mirroring configuration;

a code adapted to cause the controller to manage data storage among the at least three sites in a full mirroring relationship; and

a code adapted to cause the controller to create a virtual connection between instances of a storage group on two of the sites whereby no user data is transmitted on the virtual connection and data on the virtual connection is monitored through a third site.