Method and system for an overlay filesystem

• US 7,197,516 B1
• Filed: 07/22/2003
• Issued: 03/27/2007
• Est. Priority Date: 07/22/2002
• Status: Expired due to Fees

• Current Assignee: Veritas Technologies, LLC (Whitehouse Group Ltd.)
• Original Assignee: Veritas Operating Corporation (Gen Digital Inc.)
• Inventors: Wong, Carlos, Hipp, Burton Akira, Yeh, Yuh-Yen
• Primary Examiner(s): MIZRAHI, DIANE D
• Application Number:
  US10/625,219
• Time in Patent Office: 1,344 Days
• Field of Search: 707 1- 10, 707100-1041, 707200-206, 709/203, 717/178
• US Class Current: 1/1
• CPC Class Codes: G06F 16/10 File systems; File servers

Abstract

An overlay filesystem is described. The overlay filesystem is capable of accessing files from a first layer or a second layer based on one or more criteria. The overlay filesystem is also capable of saving changes made to any files in a different layer while preserving the original files in the original layer.

Claims

1. A system comprising:

at least one processor; and

at least one memory coupled to the at least one processor, wherein the at least one memory is configured to store;

a vnode layer configured to receive a plurality of requests for filesystem operations generated by at least one application, wherein the plurality of requests for the filesystem operations comprise an open( ) request specifying a file; and

an overlay filesystem comprising;

a back filesystem containing shared read-only files; and

a front filesystem mounted above the back filesystem and containing writable files;

wherein the overlay filesystem is configured to;

selectively route the plurality of requests for the filesystem operations from the vnode layer to the front and back filesystems; and

allocate an onode in the at least one memory upon opening the file specified by the open( ) request, wherein the onode corresponds to the file specified by the open( ) request and comprises a shadow vnode.

2. The system of claim 1, wherein the overlay filesystem is configured to:

send the open( ) request to the front filesystem which returns a front vnode to be stored in the onode, and

send the open( ) request to the back filesystem which returns a back vnode to be stored in the onode.

3. The system of claim 2,

wherein the shadow vnode maintains a reference count that is incremented each time the file is opened.

4. The system of claim 2,

wherein the onode and the shadow vnode are linked and the shadow vnode is returned to the vnode layer.

5. The system of claim 4,

wherein the vnode layer returns to the application a file descriptor linked to the shadow vnode.

6. The system of claim 1,

wherein the plurality of requests for the filesystem operations comprise a read( ) request specifying a second file;

wherein the overlay filesystem is configured to

receive the read( ) request and a second shadow vnode from the vnode layer;

determine a vnode for a filesystem from an onode for the second file; and

pass the read( ) request to the filesystem.

7. The system of claim 6,

wherein the vnode for the filesystem is a front vnode for the front filesystem.

8. The system of claim 1,

wherein a structure of the vnode layer is filesystem-independent;

and

wherein the overlay filesystem maintains a plurality of onodes.

9. The system of claim 8,

wherein each onode comprises a vnode triplet.

10. The system of claim 9,

wherein the vnode triplet includes a shadow vnode pointer, a front vnode pointer, and a back vnode pointer that point to a shadow vnode, a front vnode, and a back vnode, respectively.

11. The system of claim 8, wherein each onode is stored in a hash table.

12. The application processing network system of claim 1,

wherein the overlay filesystem includes more than two filesystems and is further configured to allocate and cache a plurality of onodes.

13. The system of claim 1,

wherein the overlay filesystem is further configured to support a snapshot/restore module.

14. The system of claim 1,

wherein the overlay filesystem is further configured to implement a file in the front filesystem with a page-level copy-on-write structure.

15. The system of claim 14,

wherein a format of the file in the front filesystem includes a header, a page map, and a page.

16. The system of claim 15,

wherein the header stores extended file attributes, file verification data, virtual size, and reserved/padding.

17. The system of claim 15,

wherein the page map comprises a bitmap indicating a location of the page in the front filesystem.

18. A computer-implemented method comprising:

receiving a plurality of requests for filesystem operations at a vnode layer, wherein the plurality of requests for filesystem operations are generated by at least one application, and wherein the plurality of requests for the filesystem operations comprise an open( ) request specifying a file;

selectively routing the plurality of requests for the filesystem operations from the vnode layer to a back filesystem and a front filesystem in an overlay filesystem, wherein the back filesystem contains shared read-only files, and wherein the front filesystem is mounted above the back filesystem and contains writable files; and

allocating an onode in memory upon opening the file specified by the open( ) request, wherein the onode corresponds to the file specified by the open( ) request and comprises a shadow vnode.

19. A computer-readable storage medium comprising program instructions, wherein the program instructions are computer-executable to perform:

receiving a plurality of requests for filesystem operations at a vnode layer, wherein the plurality of requests for filesystem operations are generated by at least one application, and wherein the plurality of requests for the filesystem operations comprise an open( ) request specifying a file;

selectively routing the plurality of requests for the filesystem operations from the vnode layer to a back filesystem and a front filesystem in an overlay filesystem, wherein the back filesystem contains shared read-only files, and wherein the front filesystem is mounted above the back filesystem and contains writable files; and

allocating an onode in memory upon opening the file specified by the open( ) request, wherein the onode corresponds to the file specified by the open( ) request and comprises a shadow vnode.