Extension of write anywhere file system layout

• US 20050246401A1
• Filed: 04/30/2004
• Published: 11/03/2005
• Est. Priority Date: 04/30/2004
• Status: Active Grant

• Current Assignee: NetApp, Inc.
• Original Assignee: NetApp, Inc.
• Inventors: English, Robert M., Corbett, Peter F., Lewis, Blake H., Edwards, John K., Hamilton, Eric
• Granted Patent: US 7,409,494 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/188 Virtual file systems
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/0665 at area level, e.g. provisioning of virtual or logical volumes
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

A file system layout apportions an underlying physical volume into one or more virtual volumes (vvols) of a storage system. The underlying physical volume is an aggregate comprising one or more groups of disks, such as RAID groups, of the storage system. The aggregate has its own physical volume block number (pvbn) space and maintains metadata, such as block allocation structures, within that pvbn space. Each vvol has its own virtual volume block number (vvbn) space and maintains metadata, such as block allocation structures, within that vvbn space. Notably, the block allocation structures of a vvol are sized to the vvol, and not to the underlying aggregate, to thereby allow operations that manage data served by the storage system (e.g., snapshot operations) to efficiently work over the vvols. The file system layout extends the file system layout of a conventional write anywhere file layout system implementation, yet maintains performance properties of the conventional implementation.

Claims

1. A system adapted to provide a file system layout of a storage system, the system comprising:

an aggregate configured with one or more groups of disks of the storage system, the aggregate having a physical volume block number (pvbn) space defining a global storage space of blocks provided by the disks, the global storage space substantially simplifying storage management of free blocks through use of a single pool of disks; and

one or more virtual volumes (vvols) contained within the aggregate, each vvol containing one or more files organized as buffer trees, the buffer trees utilizing pvbns as block pointers to facilitate efficient read performance on read paths of the files contained in the vvol.

2. The system of claim 1 wherein the buffer trees further utilize virtual volume block numbers (vvbns), in addition to the pvbns, as pvbn, vvbn block pointer pairs, wherein the vvbns provide efficient access to per-volume block allocation metadata information.

3. The system of claim 1 wherein the aggregate maintains block allocation structures for the aggregate within the pvbn space and wherein each vvol has a virtual volume block number (vvbn) space adapted to maintain block allocation structures for the vvol.

4. The system of claim 3 wherein the block allocation structures comprise an active map, a summary map and a space map.

5. The system of claim 3 further comprising a hidden metadata root directory of the aggregate that contains a subdirectory for each vvol.

6. The system of claim 5 wherein each subdirectory contains a storage label file that contains information about the vvol.

7. The system of claim 6 wherein the information about the vvol includes a name of the vvol, whether the vvol is online and identity and state information of the vvol.

8. The system of claim 5 wherein each subdirectory has a filesystem file comprising block pointers that reference various file systems embodied as the vvols.

9. The system of claim 8 wherein the filesystem file further comprises a container file associated with each vvol, the container file embodied as a file in the aggregate and containing all blocks used by a vvol.

10. The system of claim 9 wherein a block that is freed (a free block) in the container file is returned to the global storage space of the aggregate.

11. The system of claim 9 wherein the container file comprises indirect blocks configured as a container map that provides a forward mapping of vvbns of a vvol to pvbns of the aggregate.

12. The system of claim 1 further comprising an owner map associated with the aggregate, the owner map comprising a data structure having one or more entries, each entry providing a backward mapping between a pvbn of the aggregate and a virtual volume block number (vvbn) of a vvol.

13. The system of claim 12 wherein each entry of the owner map, indexed by a pvbn, comprises a vvol identifier (vvid) and a vvbn such that, for a block in the aggregate, the entry indicates which vvol owns the block and which vvbn it maps to in the vvbn space.

14. The system of claim 1 wherein the one or more groups of disks are RAID groups.

15. A method for providing a file system layout of a storage system, the method comprising the steps of:

assembling one or more groups of disks of the storage system into an aggregate, the aggregate having a physical volume block number (pvbn) space defining a global storage space provided by the disks;

apportioning the aggregate into one or more virtual volumes (vvols) of the storage system, each vvol having a virtual volume block number (vvbn) space and containing one or more files organized as buffer trees; and

utilizing pvbn, vvbn pairs as block pointers within the buffer trees, wherein the pvbns facilitate efficient read performance on read paths of the files contained in each vvol and the vvbns provide efficient access to per-volume block allocation metadata information.

16. The method of claim 15 wherein the vvbn space of each vvol facilitates efficient data management operations at the vvol.

17. The method of claim 16 further comprising the steps of:

linking the vvbn space for each vvol to the pvbn space of the aggregate; and

flexibly allocating the global storage space of the aggregate among each vvol.

18. The method of claim 17 wherein the step of utilizing comprises the step of using the pvbns as block pointers to avoid latency associated with translations from vvbns-to-pvbns when servicing file system requests at the storage system.

19. The method of claim 18 further comprising the step of maintaining block allocation structures within the vvbn space of each vvol.

20. The method of claim 19 wherein the step of maintaining comprises the step of sizing the block allocation structures to each vvol to allow operations that manage data served by the storage system to efficiently work over the vvol.

21. The method of claim 15 wherein the one or more groups of disks are RAID groups.

22. Apparatus for providing a file system layout of a storage system, the apparatus comprising:

means for assembling one or more groups of disks of the storage system into an aggregate, the aggregate having a physical volume block number (pvbn) space defining a global storage space provided by the disks;

means for apportioning the aggregate into one or more virtual volumes (vvols) of the storage system, each vvol having a virtual volume block number (vvbn) space and containing one or more files organized as buffer trees; and

means for utilizing pvbn, wbn pairs as block pointers within the buffer trees, wherein the pvbns facilitate efficient read performance on read paths of the files contained in each vvol and the vvbns provide efficient access to per-volume block allocation metadata information.

23. A computer readable medium containing executable program instructions for providing a file system layout of a storage system, the executable instructions comprising one or more program instructions for:

assembling one or more groups of disks of the storage system into an aggregate, the aggregate having a physical volume block number (pvbn) space defining a global storage space provided by the disks;

apportioning the aggregate into one or more virtual volumes (vvols) of the storage system, each vvol having a virtual volume block number (wbn) space and containing one or more files organized as buffer trees; and

utilizing pvbn, wbn pairs as block pointers within the buffer trees, wherein the pvbns facilitate efficient read performance on read paths of the files contained in each vvol and the vvbns provide efficient access to per-volume block allocation metadata information.