Mechanism for reliable update of virtual disk device mappings without corrupting data

• US 6,161,191 A
• Filed: 05/12/1998
• Issued: 12/12/2000
• Est. Priority Date: 05/12/1998
• Status: Expired due to Term

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Slaughter, Gregory L., Block, Robert J., Traversat, Bernard A., Zheng, Xiaoyan
• Primary Examiner(s): Iqbal, Nadeem
• Application Number:
  US09/076,346
• Time in Patent Office: 945 Days
• Field of Search: 714/4, 714/5, 714/7, 714/11, 714/12, 714/13, 714/25, 714/27, 714/31, 714/43, 714/42, 714/44, 714/54, 714/56, 395/200.31, 395/200.32, 395/200.33, 395/200.38, 395/200.39, 395/200.42, 711/114, 711/115
• US Class Current: 714/4.1
• CPC Class Codes: G06F 11/1423 by reconfiguration of paths
• CPC Class Codes: G06F 11/201 between storage system components
• CPC Class Codes: G06F 2201/815 Virtual

Abstract

A cluster implements a virtual disk system that provides each node of the cluster access to each storage device of the cluster. The virtual disk system provides high availability such that a storage device may be accessed and data access requests are reliably completed even in the presence of a failure. To ensure consistent mapping and file permission data among the nodes, data are stored in a highly available cluster database. Because the cluster database provides consistent data to the nodes even in the presence of a failure, each node will have consistent mapping and file permission data. A cluster transport interface is provided that establishes links between the nodes and manages the links. Messages received by the cluster transports interface are conveyed to the destination node via one or more links. The configuration of a cluster may be modified during operation. Prior to modifying the configuration, a reconfiguration procedure suspends data access requests and waits for pending data access requests to complete. The reconfiguration is performed and the mapping is modified to reflect the new configuration. The node then updates the internal representation of the mapping and resumes issuing data access requests.

Claims

1. A distributed computing system comprising:

one or more nodes coupled to a communication link and one or more storage devices coupled to said one or more nodes wherein said one or more nodes include one or more devices; and

a highly available database accessible by said one or more nodes, wherein said database provides coherent data to said one or more nodes in the presence of a failure;

wherein a mapping of said one or more devices to said one or more storage devices is stored in said highly available database;

wherein when said mapping is updated, said one or more nodes stop issuing data requests to said one or more storage devices prior to said highly available database updating said mapping, and said one or more nodes resume issuing data requests when said mapping is updated;

wherein said nodes send an acknowledge signal to said highly available database when said nodes have stopped issuing data requests.

2. The distributed computing system of claim 1 wherein said one or more nodes wait for outstanding data requests to complete prior to sending said acknowledge signal.

3. The distributed computing system of claim 2 wherein said highly available database updates said mapping after receiving said acknowledge signals.

4. The distributed computing system of claim 3 wherein said highly available database outputs a first synchronization signal to indicate a pending update of said mapping.

5. The distributed computing system of claim 4 wherein said highly available database outputs a second synchronization signal to indicate said mapping is updated.

6. The distributed computing system of claim 5 wherein said first synchronization commands and said second synchronization commands are issued to said one or more nodes concurrently.

7. The distributed computing system of claim 6 wherein said highly available database waits for an acknowledge signal from each active node prior to updating said mapping.

8. The distributed computing system of claim 7 whereby said computer system reconfigures said connections between said one or more nodes and said one or more storage devices without losing or corrupting data.

9. A method of updating a mapping of nodes to storage devices, comprising:

storing said mapping in a highly available database wherein said database is accessible by said nodes and provides coherent data to said nodes in the presence of a failure;

said database outputting an indication to said nodes that an updated mapping is pending;

said nodes suspending data requests to said storage devices;

said nodes waiting for outstanding data requests to complete;

said nodes invalidating an internal representation of said mapping;

said nodes outputting acknowledge signals to said database;

said database waiting for said acknowledge signals from active nodes;

said database updating said mapping;

said database outputting an indication to said nodes that said update is complete;

said nodes requesting an updated version of said mapping from said database; and

said nodes resuming sending said data requests to said storage devices.

10. The method of claim 9 wherein if said database does not receive an acknowledge signal from an active node, said database outputs a cancel indication to said nodes to reactivate said internal representation of said mapping.

11. The method of claim 9 wherein said indication to said nodes are synchronization signals from said database.

12. The method of claim 11 wherein said synchronization signals are issued to said one or more nodes concurrently.

13. The method of claim 12 whereby said mapping is updated without losing or corrupting data.

14. A method of updating a mapping of nodes to storage devices, comprising:

storing said mapping in a highly available database wherein said database is a accessible by said nodes and provides coherent data to said nodes in the presence of a failure;

said database outputting an indication to said nodes that an updated mapping is pending;

said nodes suspending data requests to said storage devices;

said nodes outputting acknowledge signals to said database;

said database waiting for said acknowledge signals from said active nodes;

said database updating said mapping;

said database outputting an indication to said nodes that said update is complete; and

said nodes resuming sending said data requests to said storage devices.

15. The method of claim 14 further comprising prior to said nodes outputting acknowledge signals to said database:

said nodes waiting for outstanding data requests to complete; and

said nodes invalidating an internal representation of said mapping.

16. The method of claim 15 further comprising prior to said nodes resuming sending said data requests to said storage devices said nodes requesting an updated version of said mapping from said database.

17. The method of claim 14 wherein if said database does not receive an acknowledge signal from an active node, said database outputs a cancel indication to said nodes to reactivate said internal representation of said mapping.

18. The method of claim 17 wherein said indication to said nodes is a synchronization command from said database.

19. The method of claim 18 wherein said synchronization signal is issued to said one or more nodes concurrently.

20. The method of claim 19 whereby said mapping is updated without losing or corrupting data.

21. A computer-readable storage medium comprising program instructions for updating a mapping of nodes to storage devices, wherein said program instructions execute on a plurality of nodes of a distributed computing system and said program instructions are operable to implement the steps of:

storing said mapping in a highly available database wherein said database is accessible by said nodes and provides coherent data to said nodes in the presence of a failure;

said database outputting an indication to said nodes that an updated mapping is pending;

said nodes suspending data requests to said storage devices;

said nodes waiting for outstanding data requests to complete;

said nodes invalidating an internal representation of said mapping;

said nodes outputting acknowledge signals to said database;

said database waiting for said acknowledge signals from active nodes;

said database updating said mapping;

said database outputting an indication to said nodes that said update is complete;

said nodes requesting an updated version of said mapping from said database; and

said nodes resuming sending said data requests to said storage devices.