Filesystem data integrity in a single system image environment

• US 6,122,629 A
• Filed: 04/30/1998
• Issued: 09/19/2000
• Est. Priority Date: 04/30/1998
• Status: Expired due to Term

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Compaq Computer Corporation (HP Inc.)
• Inventors: Chow, William W., Zafman, David B., Walker, Bruce J.
• Primary Examiner(s): Black, Thomas G.
• Assistant Examiner(s): TRINH, WILLIAM B
• Application Number:
  US09/070,897
• Time in Patent Office: 873 Days
• Field of Search: 707/8, 707/201, 707/202, 707/200, 711/143, 711/151
• US Class Current: 707/613
• CPC Class Codes: G06F 16/10 File systems; File servers
• CPC Class Codes: Y10S 707/99938 Concurrency, e.g. lock management in shared database
• CPC Class Codes: Y10S 707/99952 Coherency, e.g. same view to multiple users
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

A system for protection of filesystem data integrity within a computer cluster is provided. The system uses redundant data caches at client and server nodes within the computer cluster. Caching of filesystem data is controlled so that non-shared files are preferably cached at client nodes. This increases filesystem performance within the computer cluster and ensures that failures may not result in a loss of modified filesystem data without a corresponding loss to the process(es) accessing that data. Shared files are cached at the server node and a backup cache node. This protects modified filesystem data against any single node failure.

Claims

1. A method for providing data integrity level one protection to changes made to a file against the loss of a node within a computer cluster having at least one server node and at least two client nodes, the method comprising the steps of:

storing a copy of information from the file in a client cache of a first client node of the at least two client nodes, wherein the information stored in the client cache can be accessed by processes running on the first client node;

storing a copy of information from the file in a client cache of a second client node of the at least two client nodes, wherein the information stored in the client cache can be accessed by processes running on the second client node;

executing a file change to the information stored in the client cache of the first client node and storing the file change in the client cache of the first client node;

asynchronously forwarding the file change from the client cache of the first client node to the at least one server node;

performing a filesync operation, wherein the filesync operation cause the server node to update the information in the file by including the file change.

2. The method of claim 1, further comprising the steps of:

executing a file change to the information stored in the client cache of the second client node and storing the file change in the client cache of the second client node; and

asynchronously forwarding the file change from the client cache of the second client node to the at least one server node.

3. The method of claim 2 further comprising the step of:

performing a filesync operation, wherein the filesync operation cause the server node to update the information in the file by including the file change forwarded from the client cache of the second client node.

4. A method for protecting changes made to shared files against the loss of a node within a computer cluster, the method comprising the steps of:

caching, by a server node, a file change requested by a process executing on a first remote node;

synchronously forwarding, by the server node, the file change to a backup cache node, which is different from the first remote node;

caching, by the backup cache node, the file change;

performing, by the server node, a filesync operation, the filesync operation causing the server node to record the file change within a physical filesystem; and

un-caching, by the backup cache node, the file change.

5. A method for providing data integrity level two protection to changes made to a file against the loss of a node within a computer cluster having at least one server node and at least two client nodes, the method comprising the steps of:

storing a copy of information from the file in a client cache of a first client node of the at least two client nodes, wherein the information stored in the client cache can be accessed by processes running on the first client node;

storing a copy of information from the file in a client cache of a second client node of the at least two client nodes, wherein the information stored in the client cache can be accessed by processes running on the second client node;

executing a file change to the information stored in the client cache of the first client node and storing the file change in the client cache of the first client node;

synchronously forwarding the file change from the client cache of the first client node to the at least one server node;

recording, by the server node, the file change within a filesystem;

acknowledging, by the server node, receipt of the file change and completion of recording.

6. The method of providing data integrity level two protection, as claimed in claim 5, further comprising:

synchronously forwarding the file change from the server node to a backup cache node; and

caching, by the backup cache node, the file change.

7. The method of claim 5, further comprising the steps of:

executing a file change to the information stored in the client cache of the second client node and storing the file change in the client cache of the second client node; and

synchronously forwarding the file change from the client cache of the second client node to the at least one server node.

8. The method of claim 7 further comprising the step of:

performing a filesync operation, wherein the filesync operation cause the server node to update the information in the file by including the file change forwarded from the client cache of the second client node.