Power-safe disk storage apparatus, systems, and methods

• US 7,779,294 B2
• Filed: 04/15/2005
• Issued: 08/17/2010
• Est. Priority Date: 04/15/2005
• Status: Active Grant

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Corrado, Francis R., Nemiroff, Daniel
• Primary Examiner(s): Beausoliel; Robert
• Assistant Examiner(s): Lottich; Joshua P
• Application Number:
  US11/107,396
• Publication Number:
  US 20060236029A1
• Time in Patent Office: 1,950 Days
• Field of Search: 714/6, 711/114
• US Class Current: 714/6.22
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 11/1435 using file system or storage system metadata
• CPC Class Codes: G06F 2211/1019 Fast writes, i.e. signaling the host that a write is done before data is written to disk

Abstract

Apparatus, systems, methods, and articles, operate to create a log file in a non-volatile data storage medium. The log file includes a header record, a trailer record, and at least one partial parity value associated with a stripe in a redundant array of inexpensive disks (RAID) sub-system. The stripe includes a first data strip to be updated, the first data strip located on an operational first disk drive, a second data strip located on a failed disk drive, and a stripe parity strip located on an operational disk drive, among others.

Claims

1. A method, including:

creating a log file in a non-volatile data storage medium, wherein the log file comprises a header record, a trailer record, and at least one partial parity value associated with a stripe in a redundant array of inexpensive disks (RAID) sub-system, and wherein the stripe comprises a first data strip to be updated located on an operational first disk drive, a second data strip located on a failed disk drive, and a stripe parity strip located on an operational disk drive, the at least one partial parity value being the result of an exclusive OR operation performed on the first data strip and the stripe parity strip, the at least one partial parity value being determined without regard to at least one data strip in the stripe; and

performing a restoral including performing an exclusive OR operation on the at least one partial parity value obtained from the log file and on the first data strip.

2. The method of claim 1, further including:

reading the first data strip from the operational first disk drive;

reading the stripe parity strip from the RAID sub-system;

calculating the at least one partial parity value from the first data strip and the stripe parity strip;

writing a parity disk number, a data disk number, a logical block address, and a transfer byte count associated with the first data strip to the header record in the log file;

writing the at least one partial parity value to a main section of the log file; and

copying the header record to the trailer record in the log file after writing the at least one partial parity value to the main section.

3. The method of claim 2, wherein calculating the at least one partial parity value comprises performing an exclusive OR function on the first data strip and on the stripe parity strip.

4. The method of claim 1, further including:

comparing the header record to the trailer record following an unexpected sub-system reset to determine whether the reset has occurred during a period between a time of initiating update writes to the first data strip and to the stripe parity strip and a time of completing the update writes.

5. The method of claim 4, further including:

restoring a first consistency between the first data strip and the stripe parity strip using the log file following the unexpected sub-system reset.

6. The method of claim 5, further including:

reading the at least one partial parity value and a set of stripe parity strip and first data strip location values from the log file;

reading the first data strip from the operational first disk drive;

performing a first parity restoral operation on the at least one partial parity value and the first data strip to obtain a restored stripe parity strip value; and

overwriting the stripe parity strip with the restored stripe parity strip value.

7. The method of claim 6, further including:

creating the log file further providing that a second-dimensional parity strip calculated using the first data strip exists on the RAID sub-system and that the RAID sub-system comprises a RAID level 6 subsystem;

performing a second parity restoral operation on the restored stripe parity strip value and on all data strips in the stripe to create a restored second-dimensional parity strip value; and

overwriting the second-dimensional parity strip with the restored second-dimensional parity strip value.

8. The method of claim 1, wherein the stripe parity strip comprises a parity value calculated using the first data strip and additional data strips included in the stripe.

9. The method of claim 1, wherein the non-volatile data storage medium comprises a reserved area located on at least one disk drive included in the RAID sub-system.

10. The method of claim 1, further including:

aggregating writes of the at least one partial parity value to the log file to increase write throughput.

11. An article including a machine-accessible medium having associated information, wherein the information, when accessed, results in a machine performing:

creating a log file in a non-volatile data storage medium, wherein the log file comprises a header record, a trailer record, and at least one partial parity value associated with a stripe in a redundant array of inexpensive disks (RAID) sub-system, and wherein the stripe comprises a first data strip to be updated located on an operational first disk drive, a second data strip located on a failed disk drive, and a stripe parity strip located on an operational disk drive, the at least one partial parity value being the result of an exclusive OR operation performed on the first data strip and the stripe parity strip, the at least one partial parity value being determined without regard to at least one data strip in the stripe, and

performing a restoral including performing an exclusive OR operation on the at least one partial parity value obtained from the log file and on the first data strip.

12. The article of claim 11, wherein the information, when accessed, results in a machine performing:

reading the first data strip from the operational first disk drive;

reading the stripe parity strip from the RAID sub-system;

calculating the at least one partial parity value from the first data strip and the stripe parity strip;

writing a parity disk number, a data disk number, a logical block address, and a transfer byte count associated with the first data strip to a header record in the log file;

writing the at least one partial parity value to a main section of the log file; and

copying the header record to a trailer record in the log file after writing the at least one partial parity value to the main section.

13. The article of claim 12, wherein the information, when accessed, results in a machine performing:

writing a digital signature to the log file to be used to confirm that a header record from the log file is associated with the at least one partial parity value.

14. The article of claim 11, wherein the information, when accessed, results in a machine performing:

comparing the header record to the trailer record following an unexpected sub-system reset to determine whether the reset has occurred during a period between a time of initiating update writes to the first data strip and to the stripe parity strip and a time of completing the update writes.

15. The article of claim 14, wherein the information, when accessed, results in a machine performing:

reading a digital signature from the log file associated with the at least one partial parity value; and

verifying the digital signature to confirm that the header record from the log file is associated with the at least one partial parity value.

16. The article of claim 15, wherein the information, when accessed, results in a machine performing:

restoring a consistency between the first data strip and the stripe parity strip using the log file following the unexpected sub-system reset, wherein the reset occurs during the period between the time of initiating the update writes to the first data strip and to the stripe parity strip and the time of completing the update writes.

17. The article of claim 16, wherein the information, when accessed, results in a machine performing:

reading the at least one partial parity value and a set of stripe parity strip and first data strip location values from the log file;

reading the first data strip from the operational first disk drive;

performing a parity restoral operation on the at least one partial parity value and the first data strip to obtain a restored stripe parity strip value; and

overwriting the stripe parity strip with the restored stripe parity strip value.

18. An apparatus, including:

one or more processors;

a partial parity logging module executable by the one or more processors to create a log file in a non-volatile data storage medium, the log file comprising at least one partial parity value associated with a stripe in a redundant array of inexpensive disks (RAID) sub-system, the stripe comprising a first data strip to be updated located on an operational first disk drive, a second data strip located on a failed disk drive, and a stripe parity strip located on an operational disk drive, the at least one partial parity value being the result of an exclusive OR operation performed on the first data strip and the stripe parity strip, the partial parity value being determined without regard to at least one data strip in the stripe, and the partial parity logging module comprising at least one of a hardware module, a firmware module, or a software module stored on a computer readable medium; and

a parity restoral module executable by the one or more processors and coupled to the logging module to restore a consistency between the first data strip and the stripe parity strip by performing an exclusive OR operation on the at least one partial parity value obtained from the log file and on the first data strip.

19. The apparatus of claim 18, wherein the log file further comprises a parity disk number associated with the stripe parity strip, and a data disk number, a logical block address, and a transfer byte count associated with the first data strip.

20. The apparatus of claim 19, wherein the parity disk number, the data disk number, the logical block address, and the transfer byte count are located in a header section of the log file, and the at least one partial parity value is located in a main section of the log file.

21. The apparatus of claim 18, wherein the non-volatile data storage medium comprises a storage area on a disk drive.

22. The apparatus of claim 21, wherein the storage area comprises a reserved area located on at least one disk drive included in the RAID sub-system.

23. The apparatus of claim 18, further including:

a parity restoral module coupled to the logging module to restore a consistency between the first data strip and the stripe parity strip using the log file following an unexpected sub- system reset, the reset occurring during a period between a time of initiating update writes to the first data strip and to the stripe parity strip and a time of completing the update writes, the parity restoral module to restore the consistency by performing an exclusive OR operation on the partial parity values obtained from the log file and on the first data strip, and the parity restoral module comprising at least one of a hardware module, a firmware module, or a software module stored on a computer readable medium.

24. A system, including:

a partial parity logging module to create a log file in a non-volatile data storage medium, the log file comprising at least one partial parity value associated with a stripe in a redundant array of inexpensive disks (RAID) sub-system and the stripe comprising a first data strip to be updated located on an operational first disk drive, a second data strip located on a failed disk drive, and a stripe parity strip located on an operational disk drive, the at least one partial parity value being the result of an exclusive OR operation performed on the first data strip and the stripe parity strip, the partial parity value being determined without regard to at least one data strip in the stripe;

a parity restoral module coupled to the logging module to restore a first consistency between the first data strip and the stripe parity strip using the log file following an unexpected sub-system reset, the reset occurring during a period between a time of initiating update writes to the first data strip and to the stripe parity strip and a time of completing the update writes, the parity restoral module to restore the first consistency by performing an exclusive OR operation on the partial parity values obtained from the log file and on the first data strip;

a processor coupled to the logging module to provide data to write to the first data strip; and

a liquid crystal display coupled to the processor to display the data.

25. The system of claim 24, wherein the RAID sub-system comprises at least one of RAID level 3, RAID level 4, RAID level 5, RAID level 6, RAID level 30, or RAID level 50.

26. The system of claim 24, further comprising:

a second-dimensional parity strip located on an operational disk drive and calculated using the first data strip, wherein the RAID sub-system comprises a RAID level 6 subsystem, and wherein the parity restoral module is to restore a second consistency between the first data strip and the second-dimensional parity strip using a restored stripe parity strip and all data strips in the stripe.

27. The system of claim 24, wherein the non-volatile data storage medium comprises a reserved area located on at least one disk drive included in the RAID sub- system.

28. The system of claim 24, wherein the unexpected sub-system reset comprises an interruption of RAID sub-system processing without notification to the RAID sub-system prior to the interruption.