Mirror file system

• US 20010051955A1
• Filed: 03/19/2001
• Published: 12/13/2001
• Est. Priority Date: 03/17/2000
• Status: Active Grant

• Current Assignee: Twin Peaks Software Incorporated
• Original Assignee: Twin Peaks Software Incorporated
• Inventors: Wong, John P.
• Granted Patent: US 7,418,439 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/2069 Management of state, configuration or failover
• CPC Class Codes: G06F 11/2071 using a plurality of controllers
• CPC Class Codes: G06F 11/2079 Bidirectional techniques
• CPC Class Codes: G06F 16/10 File systems; File servers
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

A mirror file system (MFS) is a virtual file system that links two or more file systems together and mirrors between them in real time. When the MFS receives updated data from an application, all file systems linked by the MFS are updated in real time. The file systems linked and mirrored through the mirror file system can be a local file system connected to a physical device, or a network file system exported by a remote system on a network. The real-time mirroring mechanism provided by the MFS is transparent to user applications. The system administrator first sets up the mirroring mechanism by linking a file system to another file system on a single directory through an MFS mounting protocol. These two file systems and their files are linked together and become a mirroring pair. Both copies are owned by, and under the management of, the MFS. All access to files or directories in both file system go through the MFS. The user applications perform normal file system operation and file/directory operation system calls like open, read, write and close functions from the pathname of either file system. Most of the file operations (such as a read operation) only need to go to one file system under the MFS to get the data. Only when updates occur, such as a write operation, the MFS mechanism ensures that all data updates go to both the file systems. With this mirroring mechanism of the MFS, the files/directories in one file system are mirrored to their mirroring counterparts of another file system in real time. With the MFS technology, a standalone system is able to make multiple copies of data available to the application. In the network environment, multiple servers owning the same data copy can be distributed on the network and mirror the data to each other in real time to provide more efficient and more reliable service to their clients.

Claims

1. A virtual file system which provides mirroring and linking of two physical file systems, comprising:

means for mounting components of each of said two physical file systems in a single directory; and

a virtual file system data structure containing elements which respectively correspond to each of the mounted components, each of said elements having an application interface data structure with two associated pointers that respectively point to application interface data structures of a corresponding component in each of said two physical file systems.

2. The virtual file system of

claim 1

, wherein said application interface data structures correspond to a vnode structure.

3. The virtual file system of

claim 1

, wherein said components comprise directories and files.

4. A method for sharing files in a computer system, comprising the steps of:

mounting components of each of two physical file systems in a single directory, such that a copy of each component is stored in each of said two physical file systems;

receiving a request to perform a write operation on one of said components; and

performing said write operation on both copies of said one component in said two physical file systems, respectively, in real time in response to said request.

5. The method of

claim 4

wherein said request designates said one component, on which the write operation is to be performed, by means of a path name that is common to both of said physical file systems.

6. The method of

claim 4

wherein the steps of performing said write operation includes the steps of acquiring a lock for each copy of said one component, and inhibiting said write operation until both locks can be acquired.