System and method for information handling system redundant storage rebuild

• US 8,726,070 B2
• Filed: 09/27/2010
• Issued: 05/13/2014
• Est. Priority Date: 09/27/2010
• Status: Active Grant

• Current Assignee: Dell Products LP (Dell Technologies Inc.)
• Original Assignee: Dell Products LP (Dell Technologies Inc.)
• Inventors: Nelogal, Chandrashekar, Le, Bang Kim, Yu, Frank Widjaja
• Primary Examiner(s): Lottich, Joshua P
• Application Number:
  US12/890,917
• Publication Number:
  US 20120079317A1
• Time in Patent Office: 1,324 Days
• Field of Search: 714/6.22, 714/6.24, 714/6.32, 714/52
• US Class Current: 714/6.22
• CPC Class Codes: G06F 11/1092 Rebuilding, e.g. when physically replacing a failing disk
• CPC Class Codes: G06F 2211/1035 Keeping track, i.e. keeping track of data and parity changes

Abstract

Out-of-order reconstruction of a RAID storage device at a replacement storage device enables the replacement storage device to execute I/O for reconstructed regions during reconstruction of the replacement storage device. In one embodiment, the failed storage device is analyzed to find recoverable information, which is copied to the replacement storage device to reduce the need for reconstruction. In another embodiment, the priority for region'"'"'s reconstruction is increased upon detection of an I/O to the region. The I/O is queued until reconstruction of the region and then executed after reconstruction of that region so that I/O at the region need not be repeated during reconstruction of the remainder of the replacement storage device.

Claims

1. A method for rebuilding a first storage device of a RAID array at a replacement storage device, the RAID array having plural storage devices, each storage device storing information in logical units, each logical unit having plural stripe units, the method comprising:

establishing an initial order for rebuilding regions of the first storage device from the plural storage devices at the replacement storage device, each region having one or more stripe units;

detecting an I/O command for writing to a first region of the first storage device;

queuing the I/O command;

prioritizing re-creation of the first region at the replacement storage device by changing the initial order to move the first region up in the rebuild order; and

performing the queued I/O command at the replacement storage device after the first region is re-created.

2. The method of claim 1 further comprising:

determining regions associated with the first storage device that do not have an associated I/O command; and

performing a parity operation to sequentially re-create the regions that do not have an associated I/O command on the replacement storage device from the RAID array of plural storage devices.

3. The method of claim 1 wherein establishing an initial order comprises establishing a parity operation to sequentially rebuild the regions.

4. The method of claim 3 wherein each region rebuild retrieves plural stripes from plural RAID array storage devices.

5. The method of claim 1 rebuilding the first region by retrieving plural stripes associated with the first region from the RAID array storage devices.

6. The method of claim 5 wherein rebuilding the first region further comprises adjusting rebuild pointers to cause the first region to rebuild out of the initial order.

7. The method of claim 1 further comprising:

analyzing other queued I/O commands to determine any queued I/O commands associated with the first region; and

performing all queued I/O commands associated with the first region upon completing rebuild of the first region.

8. The method of claim 1 wherein the plural storage devices comprise hard disk drives.

9. A method for reconstructing a first storage device of a RAID array at a replacement storage device, the RAID array having plural storage devices, the method comprising:

establishing a sequential order for reconstructing plural regions of the first storage device;

detecting an I/O to a first region of the first storage device of the RAID array;

increasing the priority of reconstruction of the first region at the replacement storage device of the RAID array from the sequential order;

reconstructing the first region with the increased priority by retrieving stripes associated with the first region from the RAID array; and

performing the I/O at the first region of the replacement storage device after completion of the reconstruction of the first region at the replacement storage device.

10. The method of claim 9 wherein the I/O comprises a write of information to the first region.

11. The method of claim 9 wherein increasing the priority further comprises:

creating a queue of I/O to plural regions of the first storage device, the queue having the order in which I/O are received, each I/O associated with one or more regions of the first storage device; and

reconstructing the regions of the first storage device at the replacement storage device in the order of the queue.

12. The method of claim 11 further comprising:

determining that the queue is empty; and

reconstructing the first storage device at the replacement storage device in the sequential order of regions.

13. The method of claim 9 further comprising:

determining that the first region associated with an I/O has been reconstructed at the time of the I/O; and

executing the I/O based upon the determining.

14. The method of claim 9 wherein reconstructing the first region further comprises reconstructing the first region with a parity operation.

15. The method of claim 9 further comprising:

analyzing other queued I/O commands to determine any queued I/O commands associated with the first region; and

performing all queued I/O commands associated with the first region upon completing reconstruction of the first region.

16. An information handling system comprising:

plural components operable to process information;

plural storage devices operable to store the information;

a RAID controller interfaced with the plural storage devices to coordinate storage of the information at the storage devices as an array having redundancy;

a reconstruction module interfaced with the RAID controller and operable to reconstruct information stored on a first of the plural storage devices onto a replacement storage device, the reconstruction module reconstructing the first storage device in a predetermined order using a parity operation that retrieves information from the array; and

a priority module interfaced with the reconstruction module and operable to adjust the predetermined order based upon one or more predetermined factors, the predetermined factors including at least detection of an I/O directed to the first storage device, the priority module increasing the priority for reconstruction of a region of the first storage device that is associated with the I/O.

17. The information handling system of claim 16 wherein the predetermined factors comprise recoverable information on the first storage device, the priority module reconstructing the first storage device on the replacement storage device by copying the recoverable information from the first storage device to the replacement storage device before reconstructing other information with the rebuild operation.

18. The information handling system of claim 16 wherein the priority module is further operable to search a queue of I/O for I/O directed to a reconstructed region and to execute the I/O directed to the reconstructed region upon completing reconstruction of the reconstructed region.

19. The information handling system of claim 16 further comprising a priority table populated with I/O directed to the first storage region and an order of receipt of the I/O directed to the first storage region, the priority module operable to reconstruct the first storage device at the replacement storage device according to the order of receipt of the I/O.

20. The information handling system of claim 16 wherein the priority module is further operable to determine if reconstruction has occurred in a region of the replacement storage device and to allow writes to the region of the replacement storage device during reconstruction of the replacement storage device.