Failover architecture for a distributed storage system

• US 7,266,556 B1
• Filed: 12/29/2000
• Issued: 09/04/2007
• Est. Priority Date: 12/29/2000
• Status: Expired due to Fees

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Coates, Joshua L.
• Primary Examiner(s): HWANG, JOON H
• Application Number:
  US09/753,332
• Time in Patent Office: 2,440 Days
• Field of Search: 707 1- 10, 707200-206, 709200-203, 711161-162, 714/100, 714/6
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1662 the resynchronized component or unit being a persistent storage device re-synchronization of failed mirror storage G06F11/2082; rebuild or reconstruction of parity RAID storage G06F11/1008
• CPC Class Codes: G06F 11/2089 Redundant storage control functionality
• CPC Class Codes: G06F 11/2094 Redundant storage or storage space G06F11/2056 takes precedence
• CPC Class Codes: G06F 16/134 Distributed indices
• CPC Class Codes: G06F 16/1824 implemented using Network-attached Storage [NAS] architecture distributed or networked storage systems G06F3/067; protocols for distributed storage of data in a network H04L67/1097
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

A network storage system includes a virtual file system (“VFS”) that manages the files of the network storage system, and a storage center that stores the files. The VFS and the storage center are separated, such that a client accesses the VFS to conduct file system operations and the client accesses the storage center to upload/download files. The client accesses the network storage system through one or more storage ports. The storage center includes a plurality of distributed object storage managers (DOSMs) and a storage cluster that includes a plurality of intelligent storage nodes. The network storage system includes additional storage centers at geographically disparate locations. The network storage system uses a multi-cast protocol to maintain file information at the DOSMs regarding files stored in the intelligent storage nodes, including files stored in disparate storage centers.

Claims

1. A method comprising:

providing a plurality of distributed object storage managers “

DOSMs”

for receiving requests for files in a network storage file system;

providing at least three intelligent storage nodes directly being accessed to each of said DOSMs over a wide area, public access network coupling the DOSMs to the intelligent storage nodes, said intelligent storage nodes being accessed to said DOSMs via public access network addresses associated with said intelligent storage nodes, each intelligent storage node including a processor core and a plurality of storage devices;

storing at least one file in a first intelligent storage node accessed via a DOSM over said network;

storing a duplicate of said file in a second intelligent storage node accessed via said network;

in an event of a failure of said first intelligent storage node resulting in a failover condition rendering said first intelligent storage node unavailable, upon receiving a request for said file by said DOSM, identifying by said DOSM that said second intelligent storage node stores said duplicate of said file, redirecting said file request, via said network, to said second intelligent storage node and indicating a location determined at said DOSM for said file in said second intelligent storage node; and

accessing, via said network, said file stored in said second intelligent storage node in response to said file request or a subsequent file request, wherein

storing at least one file in a first intelligent storage node accessed via a DOSM over said network comprises accessing said first intelligent storage node via a first network address;

storing a duplicate of said file in a second intelligent storage node accessed via said network comprises accessing said second intelligent storage node via a second network address; and

determining a location for said file in said second intelligent storage node comprises generating a mapping between said first network address and said second network address, wherein said first network address and said second network address comprise internet protocol (“

IP”

) network addresses and differ only in a subnet portion of said IP network addresses.

2. The method as set forth in claim 1, wherein:

storing at least one file in a first intelligent storage node comprises storing said file in said first intelligent storage node located in a first storage center; and

storing a duplicate of said file in a second intelligent storage node comprises storing said file in said second intelligent storage node located in a second storage center, geographically distant from said first storage center.

3. The method as set forth in claim 2, further comprising:

storing a plurality of files in a plurality of intelligent storage nodes in said first storage center;

storing duplicates of said plurality of files in a plurality of intelligent storage nodes in said second storage center, so as to provide a one to one mapping between said intelligent storage nodes in said first storage center and said intelligent storage nodes in said second storage center.

4. The method as set forth in claim 1, wherein:

storing at least one file in a first intelligent storage node comprises storing said file in said first intelligent storage node located in a first storage center; and

storing a duplicate of said file in a second intelligent storage node comprises storing said file in said second intelligent storage node located in said first storage center.

5. The method as set forth in claim 1, wherein:

storing at least one file in a first intelligent storage node comprises;

storing said file in a first storage center comprising a plurality of intelligent storage nodes; and

storing said duplicate of said file in a second storage center, geographically distant from said first storage center; and

determining a location for said file in said second intelligent storage node comprises searching for said file in said second storage center after entering said failover condition.

6. The method as set forth in claim 5, wherein identifying further comprises searching for said file using a point-to-point protocol between said distributed object storage manager (DOSM) and said second intelligent storage node.

7. The method as set forth in claim 5, wherein identifying further comprises searching for said file using a multi-cast protocol.

8. The method as set forth in claim 5, identifying further comprises searching for said file using a point-to-point protocol between said distributed object storage manager (DOSM) and an intelligent storage node.

9. The method of claim 1, wherein each intelligent storage node includes volatile memory, a network interface and a file system directory and each of the plurality of storage devices is a disk drive.

10. The method of claim 1, wherein one of the intelligent storage nodes is directly accessible to one of said DOSMs through a point-to-point connection established between the one of the intelligent storage nodes and the one of said DOSMs over the wide area, public access network.

11. The method of claim 1 wherein the DOSM includes a file lookup table to identify a location of object files stored in the intelligent storage nodes.

12. A distributed storage system comprising:

a wide area, public access network;

at least three intelligent storage nodes, each intelligent storage node including a processor core and a plurality of storage devices;

a first intelligent storage node having multiple storage devices, accessed via said network and storing at least one file in a network storage file system;

a second intelligent storage node having multiple storage devices, accessed via said network with a public access network address, and storing a duplicate of said file; and

a plurality of distributed object storage managers (DOSMs) remote from said storage nodes, any one of said DOSMs directly accessing, via an interconnection over said network, said file stored in said first intelligent storage node in response to a file request, in an event of a failure of said first intelligent storage node resulting in a failover condition rendering said first intelligent storage node unavailable, upon receiving the request for said file by said DOSM, identifying by said DOSM that said second intelligent storage node stores said duplicate of said file, redirecting said file request to said second intelligent storage node, via said network, indicating a location for said file in said second intelligent storage node and directly accessing, via said interconnection over said network, said file stored in said second intelligent storage node in response to said file request or a subsequent file request, wherein

said DOSM further accessing said first intelligent storage node via a first network address and accessing said second intelligent storage node via a second network address, said DOSM further generating a mapping between said first network address and said second network address, wherein said first network address and said second network address comprise internet protocol (“

IP”

) network addresses and differ only in a subnet portion of said IP network addresses.

13. The distributed storage system as set forth in claim 12, wherein:

a first storage center comprising said first intelligent storage node; and

a second storage center, geographically distant from said first storage center comprising said second intelligent storage node.

14. The distributed storage system as set forth in claim 13, wherein:

said first storage center comprises a plurality of files stored in a plurality of intelligent storage nodes; and

said second storage center comprises duplicates of said plurality of files stored in a plurality of intelligent storage nodes, so as to provide a one to one mapping between said intelligent storage nodes in said first storage center and said intelligent storage nodes in said second storage center.

15. The distributed storage system as set forth in claim 12, wherein said first and second intelligent storage nodes reside in a single storage center.

16. The distributed storage system as set forth in claim 12, further comprising:

a first storage center comprising a plurality of intelligent storage nodes for storing said file;

a second storage center, geographically distant from said first storage center, for storing said duplicate of said file; and

said DOSM searching for said file in said second storage center after entering said failover condition.

17. The distributed storage system as set forth in claim 16,

wherein said DOSM further comprises processes searching for said file in said first storage center if said file is not located in said second storage center.

18. The distributed storage system as set forth in claim 16, wherein identifying uses a multi-cast protocol for communicating among said DOSM and said intelligent storage nodes.

19. The distributed storage system as set forth in claim 16, wherein identifying uses a point-to-point protocol for communicating between said DOSM and one of said intelligent storage nodes.

20. The distributed storage system as set forth in claim 12, wherein the processor core communicates with the storage devices using ISA protocol.

21. The distributed storage system as set forth in claim 12, wherein the processor core communicates with the storage devices using SCSI protocol.

22. A distributed storage system comprising:

a wide area, public access network;

at least three intelligent storage nodes, each intelligent storage node including a processor core and a plurality of storage devices; and

a distributed virtual file system comprising;

at least three directories;

a first directory, remote from a requesting client and from an associated intelligent storage node, accessed via said network, storing file system information associated with said intelligent storage node;

a second directory, accessed via said network, storing a duplicate of said file system information; and

at least one distributed directory manager (DDM) directly accessing, via said network, said file system information stored in said first directory in response to a file system request for a file in a network storage file system, and redirecting, in an event of a failure of said first directory resulting in a failover condition rendering said first directory unavailable, the file system request from said first directory to said second directory, via said network, indicating a location for said file system information in said second directory, and directly accessing, via said network, said file system information stored in said second directory in response to said redirected file system request, wherein

said DDM further accessing said first directory via a first network address and accessing said second directory via a second network address, said DDM further generating a mapping between said first network address and said second network address, wherein said first network address and said second network address comprise internet protocol (“

IP”

) network addresses and differ only in a subnet portion of said IP network addresses.

23. The distributed storage system as set forth in claim 22, wherein:

a first storage center comprising said first directory; and

a second storage center, geographically distant from said first storage center, comprising said second directory.

24. The distributed storage system as set forth in claim 23, wherein:

said first storage center comprises file system information stored in a plurality of directories; and

said second storage center comprises a duplicate of said file system information stored in a plurality of directories, so as to provide a one to one mapping between said directories in said first storage center and said directories in said second storage center.

25. The distributed storage system as set forth in claim 22, wherein said first and second directories reside in a single storage center.

26. The distributed storage system as set forth in claim 22, further comprising:

a first storage center comprising a plurality of directories for storing said file system information;

a second storage center, geographically distant from said first storage center, for storing a duplicate of said file system information; and

said DDM searching for said file system information in said second storage center after entering said failover condition.

27. The distributed storage system as set forth in claim 26, wherein said DDM further comprises processes searching for said file system information in said first storage center if said file system information is not located in said second storage center.

28. The distributed storage system as set forth in claim 26, wherein a multi-cast protocol is used for communicating among said DDM and said directories.

29. The distributed storage system as set forth in claim 26, wherein a point-to-point protocol is used for communicating between said DDM and one of said directories.