REMOTE VOLUME ACCESS AND MIGRATION VIA A CLUSTERED SERVER NAMESPACE

• US 20100114889A1
• Filed: 10/30/2008
• Published: 05/06/2010
• Est. Priority Date: 10/30/2008
• Status: Active Grant

• Current Assignee: NetApp, Inc.
• Original Assignee: NetApp, Inc.
• Inventors: Corbett, Peter, Rabii, Faramarz, Vassef, Hooman, Arner, Keith
• Granted Patent: US 8,078,622 B2
• Time in Patent Office: Days
• US Class Current: 707/737
• 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

Abstract

A system and method that provides users of network data storage systems with the ability to gain the advantages of a clustered storage server system, in which volumes stored on multiple server nodes are linked into a virtual global hierarchical namespace, without first having to migrate their data to the clustered storage server system. The system employs an extended virtual global hierarchical namespace that allows client systems to access, via the extended global namespace, volumes stored on the clustered storage server system and on one or more storage servers that are remote from and do not constitute a part of the clustered system. The extended global namespace can also be employed to perform migration of volume data among the multiple nodes of the clustered storage server system and the remote storage servers.

Claims

1. A method of accessing data on a volume of a remote storage server via an extended global namespace of volumes of a cluster of storage servers, comprising the steps of:

providing a global namespace of volumes of said cluster of storage servers, each of said volumes of said cluster of storage servers being linked into said global namespace via a junction corresponding to an internal mount point in said global namespace for a respective one of said volumes of said cluster of storage servers;

receiving, at one of said storage servers of said cluster of storage servers, at least one request from at least one client regarding data on at least one volume of said remote storage server, the client request including a volume identifier for the respective remote storage server volume,

wherein said global namespace is extended to include the volume of said remote storage server, the volume of said remote storage server being linked into the extended global namespace via a junction corresponding to an internal mount point in the extended global namespace for the volume of said remote storage server;

querying, by said one of said storage servers of said cluster of storage servers using the volume identifier for the remote storage server volume, at least one database to obtain location information for the remote storage server volume, said location information being associated with said internal mount point in the extended global namespace for the remote storage server volume;

locating the remote storage server volume using said location information obtained from the database; and

accessing data associated with the located remote storage server volume.

2. The method of claim 1 wherein said receiving of the request from the client regarding data on the volume of said remote storage server includes receiving a request from the client for reading data from or writing data to the remote storage server volume, and receiving a client identifier in association with the request.

3. The method of claim 2 wherein said client identifier includes one or more of an IP address and a port number associated with said client.

4. The method of claim 2 wherein said receiving of the request from the client for reading data from or writing data to the remote storage server volume includes receiving a request from the client for reading data from or writing data to a data file stored on the remote storage server volume.

5. The method of claim 4 wherein the request received from the client includes a file handle for said data file that is compliant with said cluster of storage servers.

6. The method of claim 5 further including replacing said file handle included in the request with a file handle that is compliant with said remote storage server.

7. The method of claim 2:

wherein said accessing of data associated with the located remote storage server includes generating a tunneled request for reading data from or writing data to the remote storage server volume;

wherein said tunneled request includes an encapsulated tunneling header and an encapsulated copy of the request from the client for reading data from or writing data to the remote storage server volume; and

wherein said encapsulated tunneling header includes the client identifier.

8. The method of claim 7 wherein said accessing of data associated with the located remote storage server further includes forwarding, by said one of said storage servers of said cluster of storage servers, said tunneled request to said remote storage server.

9. The method of claim 8:

wherein said accessing of data associated with the located remote storage server further includes receiving, at said one of said storage servers of said cluster of storage servers, a tunneled response from said remote storage server;

wherein said tunneled response includes an encapsulated copy of a response to the request from the client for reading data from or writing data to the remote storage server volume.

10. The method of claim 9 wherein said accessing of data associated with the located remote storage server further includes extracting, at said one of said storage servers of said cluster of storage servers, the response to the request from the client from said encapsulated copy of said response.

11. The method of claim 10:

further including generating, at said one of said storage servers of said cluster of storage servers, a response for subsequent transmission to the client;

wherein said response for subsequent transmission to the client comprises the response extracted from said encapsulated copy of said response.

12. The method of claim 11 further including transmitting, to the client by said one of said storage servers of said cluster of storage servers via a virtual interface through which the request from the client was received, said response including the extracted response, and said client identifier in association with said response.

13. The method of claim 9 wherein said receiving of the request from the client for reading data from or writing data to the remote storage server volume includes receiving a request from the client for reading data from or writing data to a data file stored on the remote storage server volume.

14. The method of claim 13 wherein said response to the request from the client includes a file handle for said data file that is compliant with said remote storage server.

15. The method of claim 14 further including replacing said file handle included in said response with a file handle that is compliant with said cluster of storage servers.

16. The method of claim 1 wherein said receiving of the request from the client regarding data on the volume of said remote storage server includes receiving, from the at least one client, at least one request for at least one lock on at least one data file stored on the at least one volume of said remote storage server, and receiving a client identifier in association with the respective client lock request.

17. The method of claim 16:

wherein said accessing of data associated with the located remote storage server includes generating at least one tunneled lock request;

wherein the tunneled lock request includes an encapsulated tunneling header and an encapsulated copy of the client lock request; and

wherein said encapsulated tunneling header includes said client identifier.

18. The method of claim 17 wherein said accessing of data associated with the located remote storage server further includes forwarding, by said one of said storage servers of said cluster of storage servers, the tunneled lock request to said remote storage server.

19. The method of claim 18:

wherein said accessing of data associated with the located remote storage server further includes receiving, at said one of said storage servers of said cluster of storage servers, at least one tunneled response from said remote storage server;

wherein the tunneled response includes an encapsulated copy of a response to the client lock request.

20. The method of claim 19 wherein said accessing of data associated with the located remote storage server further includes extracting, at said one of said storage servers of said cluster of storage servers, said response to the client lock request from said encapsulated copy of said response.

21. The method of claim 20:

further including generating, at said one of said storage servers of said cluster of storage servers, a response for subsequent transmission to said client;

wherein said response for subsequent transmission to said client comprises the response to the client lock request extracted from said encapsulated copy of said response.

22. The method of claim 21 further including transmitting, by said one of said storage servers of said cluster of storage servers to said client, said response comprising the extracted response to the client lock request, and said client identifier in association with said response.

23. The method of claim 17 wherein the tunneled lock request further includes an encapsulated opaque handle for use in a callback message from said remote storage server in the event said at least one lock is currently unavailable.

24. The method of claim 23 further including, in the event the currently unavailable lock subsequently becomes available, receiving, at said one of said storage servers of said cluster of storage servers, said callback message from said remote storage server.

25. The method of claim 24:

wherein said callback message comprises at least one tunneled callback message; and

wherein the tunneled callback message includes a copy of the opaque handle and a copy of said callback message.

26. The method of claim 25 further including generating, at said one of said storage servers of said cluster of storage servers, a file handle for said data file that is compliant with said cluster of storage servers using the opaque handle.

27. The method of claim 24 further including transmitting, by said one of said storage servers of said cluster of storage servers, said callback message to said client.

28. The method of claim 27 further including receiving, at said one of said storage servers of said cluster of storage servers, a response to said callback message from said client.

29. The method of claim 28 further including transmitting, by said one of said storage servers of said cluster of storage servers, at least one tunneled response to said callback message, wherein the tunneled response includes an encapsulated copy of said response to said callback message received from said client.

30. The method of claim 18 further including storing, on said remote storage server in a persistent monitor list, a record of an identifier of a virtual interface, referred to as a client-facing VIF, through which each client lock request was received by a respective one of said storage servers of said cluster of storage servers, and a record of an identifier of a virtual interface, referred to as a remote storage server-facing VIF, through which each tunneled lock request was forwarded by the respective one of said storage servers of said cluster of storage servers to said remote storage server.

31. The method of claim 30 further including, in the event said remote storage server is re-booted during recovery from a system crash:

querying said persistent monitor list to identify said client-facing VIF through which each client lock request was received by a respective one of said storage servers of said cluster of storage servers;

querying said persistent monitor list to identify a corresponding remote storage server-facing VIF through which the tunneled lock request was forwarded by the respective one of said storage servers of said cluster of storage servers to said remote storage server; and

for each identified client-facing VIF, receiving, at a respective one of said storage servers of said cluster of storage servers via the identified corresponding remote storage server-facing VIF through which the tunneled lock request was previously forwarded, a tunneled notify request from said remote storage server.

32. The method of claim 31 wherein said tunneled notify request includes an encapsulated tunneling header, and an encapsulated copy of a notify request.

33. The method of claim 32 further including:

extracting, at the respective one of said storage servers of said cluster of storage servers, said notify request from said encapsulated copy of the notify request; and

forwarding, by the respective one of said storage servers of said cluster of storage servers, said notify request to the respective client via the identified client-facing VIF.

34. The method of claim 31 further including, in the event the identified client-facing VIF and the identified corresponding remote storage server-facing VIF migrate to different ones of said storage servers of said cluster of storage servers, transmitting, by the respective one of said storage servers of said cluster of storage servers in response to receipt of said tunneled notify request, a tunneled lock response re-direct message to said remote storage server.

35. The method of claim 34 wherein said tunneled lock response re-direct message includes a list of remote storage server-facing VIFs associated with one or more storage servers of said cluster of storage servers to which the corresponding identified client-facing VIF possibly migrated.

36. The method of claim 2 further including mapping said client identifier to a unique port number of said one of said storage servers of said cluster of storage servers.

37. The method of claim 36 wherein said mapping of said client identifier to said unique port number includes mapping said client identifier from an identifier of a virtual interface, referred to as a client-facing VIF, through which the client request was received by said one of said storage servers of said cluster of storage servers, to an identifier of a virtual interface, referred to as a remote storage server-facing VIF, interfacing said one of said storage servers of said cluster of storage servers to said remote storage server via said unique port number.

38. The method of claim 37 further including storing, in a client map, said mapping of said client identifier to said unique port number of said one of said storage servers of said cluster of storage servers, said client map being stored in a persistent storage within said cluster of storage servers.

39. The method of claim 38 wherein said accessing of data associated with the located remote storage server includes generating a client request including said unique port number and a copy of the request from the client for reading data from or writing data to the remote storage server volume.

40. The method of claim 39 wherein said accessing of data associated with the located remote storage server further includes forwarding, by said one of said storage servers of said cluster of storage servers, said client request including said unique port number and said copy of the request from the client to said remote storage server.

41. The method of claim 40:

wherein said accessing of data associated with the located remote storage server further includes receiving, at said one of said storage servers of said cluster of storage servers, a response to the forwarded client request from said remote storage server;

wherein said response to the forwarded client request includes said unique port number of said one of said storage servers of said cluster of storage servers.

42. The method of claim 41 further including querying, using the identifier of said remote storage server-facing VIF and said unique port number of said one of said storage servers of said cluster of storage servers, said client map to obtain said client identifier.

43. The method of claim 42:

further including generating, at said one of said storage servers of said cluster of storage servers, a response for subsequent transmission to the client;

wherein said response for subsequent transmission to the client comprises said response to the forwarded client request and said client identifier obtained from said client map.

44. The method of claim 1:

wherein said receiving of said at least one request from said at least one client is performed via a first virtual interface, referred to as a client-facing VIF, on said one of said storage servers of said cluster of storage servers; and

wherein said accessing of said data associated with the located remote storage server volume is performed via a second virtual interface, referred to as a remote storage server-facing VIF, on said one of said storage servers of said cluster of storage servers.

45. The method of claim 44 further including logically pairing said client-facing VIF and said remote storage server-facing VIF on said one of said storage servers of said cluster of storage servers.

46. The method of claim 45 further including, in the event of a failure of one or more of said client-facing VIF, said remote storage server-facing VIF, and said one of said storage servers of said cluster of storage servers, migrating the logically-paired client-facing VIF and remote storage server-facing VIF together to another one of said storage servers of said cluster of storage servers.

47. The method of claim 2 wherein said accessing of data associated with the located remote storage server includes generating a re-direct directive containing said location information obtained from the database.

48. The method of claim 47 further including transmitting, by said one of said storage servers of said cluster of storage servers, said re-direct directive to said client.

49. A method of tracking lock information pertaining to one or more lock requests issued by one or more clients to access one or more corresponding data storage entities, the corresponding data storage entities being stored on one or more storage servers within a network data storage environment including a plurality of storage servers, said method comprising the steps of:

in a first receiving step, receiving, by a first storage server, at least one lock request from at least one client to access at least one corresponding data storage entity, the corresponding data storage entity being stored on a second storage server, said first storage server having an associated lock manager;

in a forwarding step, forwarding, by said first storage server, the lock request to said second storage server to obtain the requested lock on the corresponding data storage entity;

in an obtaining step, obtaining, by said first storage server using said associated lock manager, a lock on a corresponding shadow data storage entity, said corresponding shadow data storage entity being representative of the data storage entity stored on said second storage server;

in a storing step, storing, by said first storage server using said associated lock manager, lock information pertaining to said lock on said corresponding shadow data storage entity,

wherein the stored lock information includes an identifier for the client obtaining the lock on the corresponding data storage entity, and an identifier for the corresponding data storage entity;

in a second receiving step, receiving, at said first storage server, at least one notify request from said second storage server, the notify request pertaining to at least one data storage entity stored on said second storage server for which the client obtained a lock; and

in response to the receipt of the notify request;

in a consulting step, consulting the stored lock information to identify the data storage entity stored on said second storage server for which the client obtained the lock, and to identify the client that obtained the lock on the data storage entity;

in a releasing step, releasing the lock on the shadow data storage entity representative of the identified data storage entity; and

in a transmitting step, transmitting a notify request to the identified client.

50. The method of claim 49 further including:

receiving, at said first storage server, at least one notify request from the client;

consulting the stored lock information to identify at least one data storage entity stored on said second storage server for which the client obtained at least one lock; and

transmitting, by said first storage server, at least one lock release to said second storage server to release the lock obtained by the client on the identified data storage entity.

51. The method of claim 50 wherein the identified data storage entity is a data file, and further including releasing the lock obtained by the client on a per-file basis.

52. The method of claim 49 wherein said first storage server corresponds to one of a cluster of storage servers, and wherein said second storage server corresponds to a remote storage server.

53. The method of claim 52 wherein the data storage entity stored on said second storage server for which the client obtained the lock is linked into an extended global namespace of said cluster of storage servers.

54. The method of claim 49:

wherein the data storage entity stored on said second storage server is a data file having an associated file handle;

wherein the associated file handle is contained in the lock request; and

wherein said obtaining of said lock on said corresponding shadow data storage entity includes obtaining a shared lock on the shadow data storage entity using the same file handle contained in the lock request.

55. The method of claim 49 wherein said first receiving step, said forwarding step, said obtaining step, said storing step, said second receiving step, said consulting step, said releasing step, and said transmitting step are each performed by a proxy component disposed on said first storage server.

56. The method of claim 55:

wherein said first storage server corresponds to one of a cluster of storage servers;

said method further including, in the event of a failure of said first storage server, migrating said proxy component from said first storage server to another one of said cluster of storage servers.

57. The method of claim 56:

wherein said first receiving step, said forwarding step, and said second receiving step are each performed via a virtual interface on said first storage server; and

said method further including, in the event of a failure of said virtual interface, migrating said virtual interface from said first storage server to said another one of said cluster of storage servers to which said proxy component migrated.

58. A method of migrating data from a source volume of a source storage server to a destination volume of a destination storage server, for use in a network data storage system including a cluster of storage servers and at least one remote storage server, said method comprising the steps of:

beginning a migration of said data by copying a first portion of said data from said source volume of said source storage server to said destination volume of said destination storage server, one or more of said source storage server and said destination storage server corresponding to said at least one remote storage server;

in the event a predetermined remaining portion of said data has not yet been copied from said source volume to said destination volume, pausing client access to said source storage server; and

while said client access to said source storage server is paused, performing the steps of;

completing said migration of said data by copying said predetermined remaining portion of said data from said source volume to said destination volume; and

modifying a global namespace of volumes of said cluster of storage servers to allow access to the migrated data on said destination volume via the modified global namespace.

59. The method of claim 58:

wherein said source storage server corresponds to one of said cluster of storage servers;

wherein said destination storage server corresponds to the remote storage server;

wherein said source volume is linked into said global namespace via a junction corresponding to an internal mount point in said global namespace for said source volume; and

wherein said modifying of said global namespace includes modifying said global namespace so that a master data set identifier of the junction linking said source volume into said global namespace points to a data set identifier of said destination volume.

60. The method of claim 58:

wherein said source storage server corresponds to the remote storage server;

wherein said destination storage server corresponds to one of said cluster of storage servers;

wherein said source volume is linked into said global namespace via a junction corresponding to an internal mount point in said global namespace for said source volume; and

wherein said modifying of said global namespace includes modifying said global namespace so that a master data set identifier of the junction linking said source volume into said global namespace points to a data set identifier of said destination volume.

61. The method of claim 58:

wherein said source storage server corresponds to a first remote storage server;

wherein said destination storage server corresponds to a second remote storage server;

wherein said source volume is linked into said global namespace via a junction corresponding to an internal mount point in said global namespace for said source volume; and

wherein said modifying of said global namespace includes modifying said global namespace so that a master data set identifier of the junction linking said source volume into said global namespace points to a data set identifier of said destination volume.

62. A system for accessing data on a volume of a remote storage server via an extended global namespace of volumes of a cluster of storage servers, comprising:

a cluster of storage servers;

wherein each of a plurality of volumes of said cluster of storage servers is linked into a global namespace via a junction corresponding to an internal mount point in said global namespace for a respective one of said volumes of said cluster of storage servers;

at least one remote storage server;

wherein said global namespace is extended to include at least one volume of said remote storage server, the volume of said remote storage server being linked into the extended global namespace via a junction corresponding to an internal mount point in the extended global namespace for the volume of said remote storage server;

wherein at least one storage server of said cluster of storage servers is operative;

to receive at least one request from at least one client regarding data on at least one volume of said remote storage server, the client request including a volume identifier for the respective remote storage server volume;

to query, using the volume identifier for the remote storage server volume, at least one database to obtain location information for the remote storage server volume, said location information being associated with said internal mount point in the extended global namespace for the remote storage server volume;

to locate the remote storage server volume using said location information obtained from the database; and

to access data associated with the located remote storage server volume.

63. A system for tracking lock information pertaining to one or more lock requests issued by one or more clients to access one or more corresponding data storage entities, the corresponding data storage entities being stored on one or more storage servers within a network data storage environment including a plurality of storage servers, comprising:

a first storage server;

a second storage server;

wherein said first storage server has an associated lock manager; and

wherein said first storage server is operative;

to receive at least one lock request from at least one client to access at least one corresponding data storage entity, the corresponding data storage entity being stored on said second storage server;

to forward the lock request to said second storage server to obtain the requested lock on the corresponding data storage entity;

to obtain, using said associated lock manager, a lock on a corresponding shadow data storage entity, said corresponding shadow data storage entity being representative of the data storage entity stored on said second storage server;

to store, using said associated lock manager, lock information pertaining to said lock on said corresponding shadow data storage entity,

wherein the stored lock information includes an identifier for the client obtaining the lock on the corresponding data storage entity, and an identifier for the corresponding data storage entity;

to receive at least one notify request from said second storage server, the notify request pertaining to at least one data storage entity stored on said second storage server for which the client obtained a lock; and

in response to the receipt of the notify request, to consult the stored lock information to identify the data storage entity stored on said second storage server for which the client obtained the lock and to identify the client that obtained the lock on the data storage entity, to release the lock on the shadow data storage entity representative of the identified data storage entity, and to transmit a notify request to the identified client.

64. A system for migrating data from a source volume of a source storage server to a destination volume of a destination storage server, comprising:

a plurality of storage servers;

wherein said plurality of storage servers includes a cluster of storage servers and at least one remote storage server;

wherein a first one of said plurality of storage servers corresponds to a source storage server;

wherein a second one of said plurality of storage servers corresponds to a destination storage server;

wherein at least one of said plurality of storage servers is operative;

to begin a migration of data by copying a first portion of said data from a source volume of said source storage server to a destination volume of said destination storage server, one or more of said source storage server and said destination storage server corresponding to said at least one remote storage server;

in the event a predetermined remaining portion of said data has not yet been copied from said source volume to said destination volume, to pause client access to said source storage server; and

while said client access to said source storage server is paused, to complete said migration of said data by copying said predetermined remaining portion of said data from said source volume to said destination volume, and to modify a global namespace of volumes of said cluster of storage servers to allow access to the migrated data on said destination volume via the modified global namespace.