Distributed snapshot process

• US 8,037,264 B2
• Filed: 10/23/2006
• Issued: 10/11/2011
• Est. Priority Date: 01/21/2003
• Status: Active Grant

• Current Assignee: Dell Products LP (Dell Technologies Inc.)
• Original Assignee: Dell Products LP (Dell Technologies Inc.)
• Inventors: Koning, G. Paul, Goluguri, Satyanarayana R., Lee, Hsin H., Subramanian, Vasudevan, Long, Paula, Hayden, Peter C., Vekiarides, Lazarus J.
• Primary Examiner(s): RUTZ, JARED IAN
• Application Number:
  US11/585,363
• Publication Number:
  US 20070106857A1
• Time in Patent Office: 1,814 Days
• Field of Search: 711/156, 711/162, 711/161, 709/203, 707/204, 707/639, 714/6
• US Class Current: 711/162
• CPC Class Codes: G06F 11/1464 for networked environments
• CPC Class Codes: G06F 16/10 File systems; File servers
• CPC Class Codes: G06F 2201/84 Using snapshots, i.e. a logical point-in-time copy of the data

Abstract

Systems and methods for providing an efficient partitioned resource server are disclosed. The partitioned resource server includes a plurality of individual servers, and the individual servers support a storage service. The systems and methods provide archiving processes for partitioned resources. To this end, the systems and process include a snapshot process operating on a server and capable of coordinating with other snapshot processes operating on the other servers for generating state information representative of the state of the partitioned resources.

Claims

1. A system for providing a partitioned storage service, comprising:

at least two networked servers,

a plurality of resources partitioned across the servers, such that a first subset of a resource is stored on a first server and a second subset of a resource is stored on a second server, with one of the first or second servers responsible for allowing access to the resource as a whole by networked clients;

at least one of the first and second servers maintaining a routing table indicating which of the first or second servers is responsible for which subset of a resource; and

a snapshot process operating on the servers, with the snapshot process operating on the first server holding data of at least the first subset of the partitioned resources being capable of coordinating with at least one other snapshot process operating on the second server holding data of the second subset of the partitioned resources, for generating a snapshot of those portions of the partitioned resources maintained by its respective server, and for generating snapshot state information representative of the state of the partitioned resources on each respective server, the snapshot state information being representative of a state of partitioned data maintained by each respective server, including at least information concerning a file structure of such data.

2. The system according to claim 1, wherein the snapshot process operating on one of the servers coordinates with snapshot processes operating on all other servers that hold data of the subsets of the partitioned resources residing on the one server.

3. The system according to claim 2, wherein a further step of coordinating snapshot processes between servers includes a time-stamp process for time stamping a command that coordinates at least one other snapshot process.

4. The system according to claim 1, wherein the snapshot process includes a request control process for processing requests received by the respective server.

5. The system according to claim 4, wherein the request control process includes a suspend process for suspending processing of requests by the respective server.

6. The system according to claim 4, wherein the request control process includes a time-stamp process for time stamping requests received by the respective server.

7. The system according to claim 1, wherein the snapshot process further includes a process for analyzing suspended requests to determine requests received after a selected time.

8. The system according to claim 1, further including an archive process for employing the state information to create a copy of a partitioned resource.

9. The system according to claim 1, further including an archive process for employing the state information to simultaneously create copies of a plurality of the resources partitioned across the at least two servers.

10. A process for providing a partitioned storage service, comprising the steps of:

providing at least two networked servers and plurality of resources partitioned across the servers, wherein a first subset of a resource is stored on a first server and a second subset of a resource is stored on a second server, with one of the first or second servers responsible for allowing access to the resource as a whole by networked clients;

maintaining a routing table on at least one of the first and second servers, the routing table indicating which of the first or second server is responsible for which subset of a resource; and

operating a snapshot process on the servers, with the snapshot process operating on the first server holding data of at least the first subset of the partitioned resources being capable of coordinating with at least one other snapshot process operating on the second server holding data of the second subset of the partitioned resources, for generating a snapshot of those portions of the partitioned resources maintained by each respective server, and for generating snapshot state information representative of a state of partitioned data maintained by each respective server, including information concerning at least a file structure of such data.

11. The process according to claim 10, wherein coordinating snapshot processes includes coordinating snapshot processes operating on one of the servers with snapshot processes operating on all other servers that hold data of the subsets of the partitioned resources residing on the one server.

12. The process according to claim 11, wherein coordinating snapshot processes between servers includes time-stamping a command that coordinates the at least one other snapshot process.

13. The process according to claim 10, wherein operating a snapshot process includes operating a request control process for processing requests received by the respective server.

14. The process according to claim 13, wherein the request control process includes a suspend process for suspending processing of requests by the respective server.

15. The process according to claim 13, wherein the request control process includes time stamping requests received by the respective server.

16. The process according to claim 13, further including analyzing suspended requests to determine requests received after a selected time.

17. A process for generating a snapshot of a plurality of resources partitioned across at least two networked servers, comprising

receiving an administration command at a first snapshot process executing on one of the servers, said first snapshot process then directing another snapshot process executing on another of the servers holding at least a subset of the partitioned resources to generate snapshot information for the portion of the partitioned resources that the other server is responsible for

the first snapshot process suspending pending client requests and directing the other snapshot process to suspend pending client requests, and

the snapshot generated by the other snapshot process comprising a snapshot of those portions of the partitioned resources maintained by the other server, and to generate the snapshot state information representative of a state of partitioned data maintained by the other server, including at least information concerning a file structure of such data.

18. The process according to claim 17, wherein

the administration command includes a prepare command to a second server supporting the resource for which a snapshot is being created.

19. The process according to claim 17, further comprising processing the state information to generate an archive copy of the resource.

20. The process according to claim 17, further comprising having the first and second snapshot processes release pending requests after generating the state information.

21. A storage area network, comprising

a data network having at least two servers,

a plurality of resources partitioned across the servers, such that a first subset of a resource is stored on a first server and a second subset of a resource is stored on a second server, with one of the first or second servers responsible for allowing access to the resource as a whole by networked clients,

at least one of the first and second servers maintaining a routing table indicating which of the first or second servers is responsible for which subset of a resource, and

a snapshot process operating on respective ones of the at least two servers, with the snapshot process operating on a first one of the servers holding data of at least the first subset of the partitioned resources being capable of coordinating with at least one other snapshot process operating on a second one of the servers holding data of the second subset of the partitioned resources, for generating a snapshot of those portions of the partitioned resources maintained by each respective server, and for generating snapshot state information representative of a state of partitioned data maintained by each respective server, including information concerning at least a file structure of such data.