Distributed system resource protection via arbitration and ownership

• US 20030065782A1
• Filed: 09/28/2001
• Published: 04/03/2003
• Est. Priority Date: 09/28/2001
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Wieland, Peter William, Nishanov, Gor
• Granted Patent: US 7,277,952 B2
• Time in Patent Office: Days
• US Class Current: 709/226
• CPC Class Codes: G06F 2209/5014 Reservation
• CPC Class Codes: G06F 9/5011 the resources being hardware resources other than CPUs, Servers and Terminals
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]

Abstract

In a distributed system, a resource such as a storage device is protected by an owner node'"'"'s exclusive access to it, wherein exclusive access is established via a persistent reservation on the resource. A persistent reservation is never removed, however the owner node'"'"'s persistent reservation may be challenged in an arbitration process, to take it away from a failed owner node. A challenger node challenges by changing an owner'"'"'s access key (that establishes the persistent reservation) to a challenge key, delaying, and determining whether the owner restored the owner'"'"'s access key during the delay, because a properly operating owner replaces a challenge key with its owner access key. If the owner fails to restore its owner access key, the challenger node becomes the new resource owner by replacing the challenge key with an owner access key of the challenger node. The key may include additional information to provide extensibility.

Claims

1. In a distributed computer system having nodes capable of accessing a resource, a method comprising:

at a first node, taking ownership of the resource including placing a persistent reservation on the resource that prevents other nodes from accessing the resource;

at a second node, challenging the persistent reservation by;

1) modifying the persistent reservation into a modified persistent reservation that prevents other nodes from accessing the resource, the modifying of the persistent reservation performed such that the resource cannot be accessed during the transition to the modified persistent reservation; and

2) enabling the first node to defend its persistent reservation; and

at the second node, determining whether the first node failed to defend its persistent reservation, and if not defended, taking ownership of the resource.

2. The method of claim 1 wherein the resource is a SCSI device, and wherein placing a persistent reservation on the resource includes issuing a reserve command.

3. The method of claim 1 wherein placing a persistent reservation on the resource includes providing a reservation key.

4. The method of claim 3 wherein modifying the persistent reservation into a modified persistent reservation includes modifying the reservation key.

5. The method of claim 4 wherein modifying the reservation key includes providing a new key.

6. The method of claim 4 wherein modifying the reservation key includes varying the key with a unique identifier for the second node.

7. The method of claim 6 wherein varying the key includes writing the unique identifier for the second node into a key field.

8. The method of claim 4 wherein the resource is a SCSI device, and wherein modifying the reservation key includes issuing a preempt command.

9. The method of claim 4 wherein enabling the first node to defend its persistent reservation includes delaying at the second node.

10. The method of claim 4 wherein determining whether the first node failed to defend its persistent reservation includes reading the reservation key.

11. The method of claim 4 wherein assuming ownership of the resource at the second node includes providing a new reservation key.

12. The method of claim 4 wherein providing a new reservation key includes writing a unique identifier for the second node into a key field.

13. The method of claim 1 further comprising, defending the persistent reservation at the first node, including changing the modified persistent reservation.

14. The method of claim 13 wherein the persistent reservation includes identification data that is modified in the modified persistent reservation, and wherein changing the modified persistent reservation includes restoring the identification data to its state prior to modification.

15. The method of claim 1 wherein the second node determines that no other node is challenging for ownership before challenging for ownership at the second node.

16. The method of claim 1 wherein challenging for ownership at the second node includes submitting a bid in association with the challenge.

17. The method of claim 1 further comprising, challenging the persistent reservation at a third node.

18. A computer-readable medium having computer-executable instructions for performing the method of claim 1.

19. In a network of distributed computing devices, a system comprising:

a resource arranged to be owned based on reservation data associated therewith;

a first node connected to the resource and having ownership of the resource via first reservation data, and having a first process therein that is configured to defend ownership of the resource while the first node is operational by restoring any changes to the reservation data back to the first reservation data;

a second node connected to the resource, the second node having a second process configured to challenge for ownership of the resource by modification of the reservation data into second reservation data, the modification of the reservation data being such that the resource cannot be accessed during the transition to the second reservation data;

wherein the second node modifies the first reservation data into the second reservation data, delays and then reads the reservation data to challenge for ownership, the second node obtaining ownership if it reads the second reservation data; and

wherein when properly operating, the first node defends against the challenge by restoring the second reservation data back to the first reservation data during the second node'"'"'s delay such that the second node will not read the second reservation data.

20. The system of claim 19 wherein the resource comprises a SCSI device, the reservation data comprises a SCSI reservation key, and wherein the first node and the second node read and write the reservation data via SCSI commands.

21. The system of claim 19 wherein the resource is a disk drive.

22. The system of claim 19 wherein the first and second nodes are configured to operate in a server cluster.

23. The system of claim 19 further comprising a third node connected to the resource, the third node having a third process configured to challenge for ownership of the resource by modification of the reservation data into third reservation data.

24. A computer-readable medium having stored thereon a data structure, comprising,

a reserver data field containing a reserver identifier;

a owner data field containing an owner identifier; and

the data structure associated with a resource and used to establish ownership of the resource and arbitrate for ownership of the resource, ownership being established by having the reserver identifier in the reserver data field equal to the owner identifier in the owner data field, and ownership being arbitrated for, including;

an owner, when operational, reading the reserver data field to see if it equals the owner identifier, and if not, writing the owner identifier into the reserver data field to win the arbitration and establish ownership; and

a challenger writing a challenger identifier into the reserver data field, delaying and then reading the reserver data field to see if it equals the challenger identifier, and if so, the challenger winning the arbitration and establishing ownership by writing the challenger identifier into the owner data field.

25. The data structure of claim 24 further comprising a version field having information therein that corresponds to an arrangement of the data structure fields.

26. The data structure of claim 24 further comprising a bid data field having information therein corresponding to a relative strength value.

27. The data structure of claim 24 wherein the owner and challenger are nodes, and the owner node shares access of the resource with other member nodes in a group, and wherein the data structure further comprises, a member generation data field having information therein that is updated when a former member node leaves the group to prevent further access by the former member.

28. The data structure of claim 24 wherein the first and second fields are incorporated into a reservation key.

29. The data structure of claim 24 further comprising another challenger writing another challenger identifier into the reserver data field, delaying and then reading the reserver data field to see if it equals the other challenger identifier, and if so, the other challenger winning the arbitration and establishing ownership by writing the other challenger identifier into the owner data field.

30. In a distributed computer system having nodes capable of accessing a storage device, a method comprising:

at a first node, placing a persistent reservation on the storage device including writing a first node identifier into reserver and owner fields of a reservation key that determines access to the device, accessing the device via the reservation key, and defending against challenges by writing the first node identifier into the reserver field when the first node detects that the first node identifier is in the owner field but not in the reserver field;

at a second node, challenging the persistent reservation key by writing a second node identifier into the reserver field of the reservation key, delaying, and determining whether the first node has failed to replace the second node identifier in the reserver field during the delay, and if not replaced, taking ownership of the storage device including writing the second node identifier into the owner field.

31. The method of claim 30 further comprising, at a third node, challenging the persistent reservation key by writing a third node identifier into the reserver field of the reservation key, delaying, and determining whether the first node has failed to replace the third node identifier in the reserver field during the delay, and if not replaced, taking ownership of the storage device including writing the third node identifier into the owner field.