Method, system, apparatus, and computer-readable medium for improving disk array performance

• US 7,991,969 B1
• Filed: 09/19/2010
• Issued: 08/02/2011
• Est. Priority Date: 06/10/2005
• Status: Active Grant

• Current Assignee: AmZetta Technologies, LLC (American Megatrends International, LLC)
• Original Assignee: American Megatrends Incorporated
• Inventors: Muthirisavenugopal, Vijayarankan, Chatterjee, Paresh, Narayanan, Ajit
• Primary Examiner(s): Elmore; Stephen C
• Application Number:
  US12/885,534
• Time in Patent Office: 317 Days
• Field of Search: 711/156, 711/112, 711/114, 711/170, 711/221, 714/5, 714/6, 714/7, 707/781, 707/802, 707/812
• US Class Current: 711/156
• CPC Class Codes: G06F 11/1092 Rebuilding, e.g. when physically replacing a failing disk
• CPC Class Codes: G06F 11/1662 the resynchronized component or unit being a persistent storage device re-synchronization of failed mirror storage G06F11/2082; rebuild or reconstruction of parity RAID storage G06F11/1008
• CPC Class Codes: G06F 11/2094 Redundant storage or storage space G06F11/2056 takes precedence
• CPC Class Codes: G06F 12/0868 Data transfer between cache memory and other subsystems, e.g. storage devices or host systems
• CPC Class Codes: G06F 2211/104 Metadata, i.e. metadata associated with RAID systems with parity
• CPC Class Codes: G06F 2211/1059 Parity-single bit-RAID5, i.e. RAID 5 implementations
• CPC Class Codes: G06F 2212/262 configured as RAID
• CPC Class Codes: G06F 3/0605 by facilitating the interaction with a user or administrator
• CPC Class Codes: G06F 3/0632 by initialisation or re-initialisation of storage systems
• 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 method, system, apparatus, and computer-readable medium are provided for improved maintenance of metadata relating to a mass storage array. The metadata may comprise the data structures utilized by a thin provisioning system. When the metadata changes, such as in response to the modification of the underlying data, changed metadata is created in the memory. A parity block is then read from the row of the array where the changed metadata is to be stored. A new parity is calculated for the row using only the old metadata, the changed metadata, and the parity read from the stripe. The old metadata need not be read from disk expressly, as is usually done. Instead, the value that is present in memory before the metadata change is utilized directly to calculate the new parity. The changed metadata and the new parity are then written to the array.

Claims

1. A method for writing metadata to a mass storage array comprising a plurality of mass storage devices, the array being divided into stripes having at least one block written to each of the mass storage devices, the method comprising:

maintaining metadata relating to the array in a structure on the array;

detecting a change to existing metadata stored in the structure;

reading a parity block from a stripe of the array where changed metadata is to be stored in the structure;

calculating a new parity for the stripe using only the existing metadata, the changed metadata, and the parity read from the array;

storing the changed metadata in the structure on the array; and

storing the new parity in the structure on the array.

2. The method of claim 1, further comprising providing a thin provisioning module that maintains details of the mass storage array.

3. The method of claim 2, wherein the details of the mass storage array are stored in a second memory.

4. The method of claim 3, wherein the second memory is a random access memory.

5. The method of claim 2, the thin provisioning module is provided as a kernel module.

6. The method of claim 1, wherein the method is performed without reading the existing metadata from the structure on the array.

7. The method of claim 1, wherein the method is performed each time a new write is made to the mass storage array.

8. A system for writing metadata relating to a mass storage array to a data structure on the mass storage array, comprising:

a plurality of mass storage devices;

a storage controller having a management layer, the management layer creating the mass storage array from the plurality of mass storage devices; and

a kernel module having a thin provisioning client that allocates physical capacity on the mass storage devices on an as-needed basis,

wherein the mass storage array is divided into stripes having at least one block written to each of the mass storage devices, wherein the thin provisioning client maintains the metadata in the data structure on the array, and

wherein when a change to the existing metadata stored in the structure is detected, a parity block is read from a stripe of the array where changed metadata is to be stored, and a parity for the stripe is calculated using only the existing metadata, the changed metadata, and the parity read from the array, and the changed metadata and new parity is stored in the structure on the array.

9. The system of claim 8, wherein the data structure having details of the mass storage array are stored in a second memory.

10. The system of claim 9, wherein the second memory is a random access memory.

11. The system of claim 9, wherein the existing metadata is not read from the structure on the array.

12. The system of claim 11, wherein the changed metadata and new parity is stored in the structure on the array is stored each time a new write is made to the mass storage array.

13. A tangible computer readable medium comprising computer executable instructions for writing metadata to a mass storage array comprising a plurality of mass storage devices, the array being divided into stripes having at least one block written to each of the mass storage devices, the instructions when executed by a computing device causing the computing device to perform the method, comprising:

maintaining metadata relating to the array in a structure on the array;

detecting a change to existing metadata stored in the structure;

reading a parity block from a stripe of the array where changed metadata is to be stored in the structure;

calculating a new parity for the stripe using only the existing metadata, the changed metadata, and the parity read from the array;

storing the changed metadata in the structure on the array; and

storing the new parity in the structure on the array.

14. The tangible computer readable medium of claim 13, further comprising instructions for providing a thin provisioning module that maintains details of the mass storage array.

15. The tangible computer readable medium of claim 14, wherein the details of the mass storage array are stored in a second memory.

16. The tangible computer readable medium of claim 15, wherein the second memory is a random access memory.

17. The tangible computer readable medium of claim 14, wherein the thin provisioning module is provided as a kernel module.

18. The tangible computer readable medium of claim 13, wherein the method is performed without reading the existing metadata from the structure on the array.

19. The tangible computer readable medium of claim 13, wherein the method is performed each time a new write is made to the mass storage array.