System And Method For Managing Replication In An Object Storage System

• US 20120290535A1
• Filed: 05/12/2011
• Published: 11/15/2012
• Est. Priority Date: 05/12/2011
• Status: Active Grant

• Current Assignee: Dell Products LP (Dell Technologies Inc.)
• Original Assignee: Dell Products LP (Dell Technologies Inc.)
• Inventors: Khosrowpour, Farzad, Patel, Dharmesh
• Granted Patent: US 8,433,681 B2
• Time in Patent Office: Days
• US Class Current: 707/634
• CPC Class Codes: G06F 11/006 Identification G06F11/2289 takes precedence
• CPC Class Codes: G06F 11/2094 Redundant storage or storage space G06F11/2056 takes precedence
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: G06F 3/065 Replication mechanisms

Abstract

An object storage system, such as a content addressed storage system, manages replication of objects across network locations to balance storage space and data security. Network locations set a policy of replicating each object at the object'"'"'s primary network location and a secondary network location. The secondary network location creates a first replica of the object and a virtual unique identifier representing a second replica of the object at the secondary network location. Creation of the second replica is suppressed unless the first replica becomes invalid so that storage space is conserved without substantially increasing the risk of loss of information represented by the object.

Claims

1. A method for storing information at plural network locations, the method comprising:

storing information at a primary network location;

associating the information with a unique identifier;

storing a replica of the information at a secondary network location;

associating the replica with a unique identifier;

generating a virtual unique identifier at the secondary network location;

detecting a failure associated with the replica;

in response to the detecting, re-generating the replica; and

associating the re-generated replica with the virtual unique identifier.

2. The method of claim 1 wherein the first and second network locations comprise an object storage system having content addressed storage.

3. The method of claim 1 wherein detecting a failure further comprises:

requesting the unique identifier;

responding to the requesting with a low priority by the virtual unique identifier relative to the unique identifier;

attempting without success to respond to the requesting with the unique identifier; and

responding to the requesting with the virtual unique identifier based upon the low priority.

4. The method of claim 1 wherein detecting a failure comprises:

attempting to access the replica at the secondary network location; and

failing to retrieve the replica in response to the attempting.

5. The method of claim 4 wherein re-generating the replica comprises:

retrieving the information from the primary network location to the secondary network location; and

storing the information at the secondary network location in association with the virtual unique identifier.

6. The method of claim 5 further comprising assigning an increased priority to retrieving the information from the primary network in response to the failing to retrieve the replica.

7. The method of claim 1 wherein detecting a failure comprises:

monitoring the secondary network location for the presence of the replica unique identifier; and

re-generating the replica if the replica unique identifier is not present.

8. The method of claim 7 further comprising assigning an increased priority to re-generating the replica if the replica unique identifier is not present.

9. The method of claim 1 wherein generating a virtual unique identifier further comprises:

detecting storage of the replica at the secondary network location;

overriding a setting to create a copy of the replica at the secondary network; and

providing the virtual unique identifier to indicate that the copy of the replica was created.

10. A system for managing replication in an object storage system having plural network locations, the system comprising:

a publisher module running on a processor at a first network location and operable to store an object at the first network location in association with a unique identifier; and

a subscriber module running on a processor at a second network location interfaced with the first network location, the subscriber module operable to create a first replica of the object at the second network location in association with a unique identifier and to create a virtual unique identifier at the second network location, the virtual unique identifier indicating a second replica of the object stored at the second network location without creation of the second replica.

11. The system of claim 10 further comprising a replica policy engine interfaced with the subscriber module, the replica policy engine operable to detect an invalid first replica and to create the second replica in response to detecting the invalid first replica.

12. The system of claim 11 wherein the replica policy engine detects the invalid first replica by monitoring the status of the first replica.

13. The system of claim 11 wherein the replica policy engine detects the invalid first replica in response to a failed attempt by an application to retrieve the first replica.

14. The system of claim 11 wherein the replica policy engine is further operable to create a second virtual unique identifier upon creating the second replica, the second virtual unique identifier associated with a third replica without creation of the third replica.

15. The system of claim 10 wherein the first and second network locations comprise content addressed storage.

16. A method for managing replication of objects at a network location of a content addressed storage system, the method comprising:

setting the network location to create at least one replica object of each object stored at the network location, each object and replica object having a unique identifier;

detecting that an object stored at the network location is a replica of an object stored at a different network location of the content addressed storage system; and

in response to the detecting, creating a unique identifier for a replica of the object stored at the network location and suppressing creation of a replica of the object stored at the network location.

17. The method of claim 16 further comprising:

determining that the object stored on the network location is not valid; and

in response to the determining, creating the replica.

18. The method of claim 17 wherein determining further comprises detecting a failed attempt to retrieve the object stored on the network location.

19. The method of claim 17 wherein determining further comprises monitoring the status of the object stored on the network location to detect that the object stored on the network location is not valid.

20. The method of claim 17 wherein in response to the determining, creating the replica further comprises assigning an increased priority to the creating the replica.