Network file system record lock recovery in a highly available environment

• US 7,356,531 B1
• Filed: 07/25/2003
• Issued: 04/08/2008
• Est. Priority Date: 07/25/2003
• Status: Active Grant

• Current Assignee: Veritas Technologies, LLC (Whitehouse Group Ltd.)
• Original Assignee: Symantec Operating Corporation (NortonLifeLock Inc.)
• Inventors: Agarwal, Manish M., Popelka, Paul L.
• Primary Examiner(s): LE, UYEN T
• Application Number:
  US10/627,258
• Time in Patent Office: 1,719 Days
• Field of Search: 707/8, 707/9, 707/200, 707/201, 707/202, 709/225, 714/1, 714/2, 714/3, 714/4, 710/200
• US Class Current: 707/682
• CPC Class Codes: G06F 11/2028 eliminating a faulty processor or activating a spare
• CPC Class Codes: G06F 11/203 using migration
• CPC Class Codes: G06F 16/10 File systems; File servers
• CPC Class Codes: Y10S 707/99938 Concurrency, e.g. lock management in shared database

Abstract

A computer accessible medium may be encoded with a plurality of instructions which, when executed in a first node of a plurality of nodes in response to a fail over of a first service group of a plurality of service groups from a second node of the plurality of nodes, initiate lock recovery for locks on one or more filesystems included in the service group. Locks on one or more filesystems in a second service group of the plurality of service groups may be maintained during a time period that locks in the first service group are recovered. The lock recovery may be initiated using a first client list of a plurality of client lists, wherein the first client list is included in the first service group and identifies clients having at least one lock on a filesystem included in the first service group. During the fail over of a service group, lock services may not be interrupted for other service groups on any of the nodes.

Claims

1. A cluster comprising:

a plurality of filesystems, each of the plurality of filesystems included in one of a plurality of service groups, and each of the plurality of service groups including one or more filesystems of the plurality of filesystems; and

a plurality of nodes, each of the plurality of nodes comprising a computer system, wherein the plurality of nodes are configured to;

(i) act as a server for the plurality of filesystems;

(ii) provide record locking services in the plurality of filesystems; and

(iii) maintain a plurality of client lists, each of the plurality of client lists included in a respective service group of the plurality of service groups and identifying clients having at least one lock on one of the one or more filesystems included in the respective service group, wherein a separate client list of the plurality of client lists is included in each respective service group of the plurality of service groups.

2. The cluster as recited in claim 1 wherein each of the plurality of client lists is stored in at least one of the one or more filesystems included in the respective service group.

3. The cluster as recited in claim 1 wherein, in response to detecting a failure with respect to a first service group on a first node of the plurality of nodes, a second node of the plurality of nodes is configured to begin acting as the server for the one or more filesystems in the first service group, and wherein the second node is configured to initiate lock recovery for locks in the one or more filesystems included in the first service group responsive to detecting the failure.

4. The cluster as recited in claim 3 wherein the second node is configured to read a first client list of the plurality of client lists, the first client list included in the first service group, to initiate lock recovery.

5. The cluster as recited in claim 3 wherein the second node is configured to initiate lock recovery by notifying each of the clients in the first client list that the clients should reclaim locks previously granted on a filesystem within the first service group.

6. The cluster as recited in claim 5 wherein the second node is configured to notify each of the clients by transmitting one or more server identifiers within the first service group to the clients.

7. The cluster as recited in claim 6 wherein each of the one or more server identifiers comprises an internet protocol address.

8. The cluster as recited in claim 5 wherein the second node is configured to initiate a period of time for the clients in the first client list to reclaim locks, wherein the second node is configured not to grant new locks on filesystems within the first service group during the period.

9. The cluster as recited in claim 8 wherein, if the second node is also acting as a server for one or more filesystems in a second service group of the plurality of service groups, the second node is configured to grant a lock requested by a client in a filesystem within the second service group during the period.

10. The cluster as recited in claim 3 wherein, if the second node is also acting as a server for one or more filesystems in a second service group of the plurality of service groups, the second node is configured to not initiate lock recovery for locks in the second service group responsive to the fail over of the first service group.

11. A method comprising:

maintaining a plurality of client lists, each of the plurality of client lists included in a respective service group of a plurality of service groups and identifying clients having at least one lock on at least one filesystem included in the respective service group, wherein a separate client list of the plurality of client lists is included in each respective service group of the plurality of service groups;

detecting a failure with respect to a first service group on a first node of the plurality of nodes;

responsive to detecting the failure, a second node of the plurality of nodes beginning to act as a server for the one or more filesystems in the first service group; and

the second node initiating lock recovery for locks on one or more filesystems including in the first service group responsive to the detecting using a first client list of the plurality of client lists, the first client list included in the first service group.

12. The method as recited in claim 11 further comprising storing each of the plurality of client lists in at least one of the one or more filesystems included in the respective service group.

13. The method as recited in claim 11 wherein initiating lock recovery comprises notifying each of the clients in the first client list that the clients should reclaim locks previously granted on at least one filesystem in the first service group.

14. The method as recited in claim 13 wherein notifying each of the clients comprises transmitting one or more server identifiers included in the first service group.

15. The method as recited in claim 13 further comprising:

initiating a period of time for the clients in the first client list to reclaim locks in the filesystems included in the first service group; and

the second node not granting new locks in the filesystems included in the first service group during the period.

16. The method as recited in claim 15 wherein the second node is also acting as a server for one or more filesystems in a second service group of the plurality of service groups, the method further comprising the second node not interrupting locking services for the second service group during the period.

17. The method as recited in claim 16 wherein not interrupting locking services comprises granting a lock in a filesystem within the second service group during the period.

18. The method as recited in claim 15 wherein the first node is also acting as a server for one or more filesystems in a second service group of the plurality of service groups, the method further comprising the first node not interrupting locking services for the second service group during the period.

19. The method as recited in claim 11 wherein the second node is also acting as a server for one or more filesystems in a second service group of the plurality of service groups, the method further comprising the second node not interrupting locking services for the second service group during the period.

20. A computer accessible medium encoded with a plurality of instructions which, when executed in a first node of a plurality of nodes in response to detecting a failure with respect to a first service group of a plurality of service groups in a second node of the plurality of nodes and each of the plurality of service groups including at least one filesystem:

initiate lock recovery for locks on each filesystem in the first service group using a first client list of a plurality of client lists, wherein each of the plurality of client lists is included in a respective service group of the plurality of service groups and identifies clients having at least one lock in at least one filesystem in the respective service group, and wherein the first client list is included in the first service group, wherein a separate client list of the plurality of client lists is included in each respective service group of the plurality of service groups.

21. The computer accessible medium as recited in claim 20 wherein each of the plurality of client lists is stored in the respective service group.

22. The computer accessible medium as recited in claim 20 wherein the plurality of instructions, when executed, initiate lock recovery by notifying each of the clients in the first client list that the clients should reclaim locks previously granted on a filesystem in the first service group.

23. The computer accessible medium as recited in claim 22 wherein notifying each of the clients in the first client lists comprises transmitting one or more server identifiers included in the first service group.

24. The computer accessible medium as recited in claim 22 wherein the plurality of instructions, when executed:

initiate a period of time for the clients in the first client list to reclaim locks on the at least one filesystem in the first service group; and

do not grant new locks on the at least one filesystem in the first service group during the period.

25. The computer accessible medium as recited in claim 24 wherein, if the second node is also acting as a server for one or more filesystems in a second service group of the plurality of service groups, the plurality of instructions, when executed, grant a lock requested by a client on a filesystem in the second service group during the period.

26. The computer accessible medium as recited in claim 20 wherein, if the second node is also acting as a server for one or more filesystems in a second service group of the plurality of service groups, the plurality of instructions, when executed, do not initiate lock recovery for locks on filesystems in the second service group responsive to the fail over of the first service group.

27. A computer accessible medium encoded with a plurality of instructions which, when executed in a first node of a plurality of nodes in response to detecting a failure with respect to a first service group of a plurality of service groups in a second node of the plurality of nodes and each of the plurality of service groups comprises at least one filesystem:

initiate lock recovery for first locks on at least one first filesystem in the first service group, the lock recovery initiated by the first node which is beginning to act as a server for the at least one first filesystem, and wherein the second node was previously acting as the server for the at least one filesystem; and

maintain second locks in at least one second filesystem in a second service group of the plurality of service groups during a time period that the first locks in the first service group are recovered.

28. The computer accessible medium as recited in claim 27 wherein the first node is acting as a server for the at least one second filesystem in the second service group.

29. The computer accessible medium as recited in claim 27 wherein a different node of the plurality of nodes is acting as the server for the at least one second filesystem in the second service group.

30. The computer accessible medium as recited in claim 27 wherein the plurality of instructions, when executed:

initiate a period of time for clients to reclaim locks on the at least one first filesystem in the first service group; and

do not grant new locks on the at least one first filesystem in the first service group during the period.

31. The computer accessible medium as recited in claim 30 wherein the plurality of instructions, when executed, grant a lock requested by a client on one of the at least one second filesystem included in the second service group during the period.

32. The computer accessible medium as recited in claim 27 wherein the plurality of instructions, when executed, do not interrupt locking services for the second service group.

33. A method comprising:

initiating lock recovery for first locks on one or more first filesystems in a first service group of a plurality of service groups by a first node of a plurality of nodes that is beginning to act as a server the one or more filesystems in response to detecting a failure with respect to the first service group in a second node of the plurality of nodes that was previously acting as the server for the one or more first filesystems; and

maintaining second locks on one or more second filesystems in a second service group of the plurality of service groups during a time period that the first locks in the first service group are recovered.

34. The method as recited in claim 33 wherein the first node is acting as a server for the one or more second filesystems in the second service group.

35. The method as recited in claim 33 wherein a different node of the plurality of nodes is acting as the server for the one or more second filesystems in the second service group.

36. The method as recited in claim 33 further comprising:

initiating a period of time for clients to reclaim locks on the one or more first filesystems in the first service group; and

not granting new locks on the one or more first filesystems in the first service group during the period.

37. The method as recited in claim 36 further comprising granting a lock requested by a client on a filesystem in the second service group during the period.