Frequently-redundant array of independent disks

• US 5,720,025 A
• Filed: 01/18/1996
• Issued: 02/17/1998
• Est. Priority Date: 01/18/1996
• Status: Expired due to Term

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Company (HP Inc.)
• Inventors: Savage, Stefan, Wilkes, Andrew John
• Primary Examiner(s): Beausoliel, Jr., Robert W.
• Assistant Examiner(s): Baderman, Scott T.
• Application Number:
  US08/588,140
• Time in Patent Office: 761 Days
• Field of Search: 395/182.04, 395/182.05, 395/441
• US Class Current: 714/6.12
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 2211/1009 Cache, i.e. caches used in RAID system with parity
• CPC Class Codes: G06F 2211/1059 Parity-single bit-RAID5, i.e. RAID 5 implementations

Abstract

There is provided a method and apparatus for effectuating data updates on a fault-tolerant storage device having a plurality of disk-drive units forming a disk array, with parity-block updates temporarily suspended to provide increased performance. Parity-block updates are typically deferred until the array is idle. Thus, data is written onto the array with no parity error correction information being present for a short amount of time. Data updates are effectuated by identifying the blocks in the array where data is to be written. New data is then written directly to the identified blocks without reading the old data present therein. A "needs parity rebuild bit" (NPRB) is written to non-volatile memory, identifying the stripes not having parity information on the array. The number of NPRBs present is monitored to allow continuous calculation of a mean-time-to-data-loss (MTTDL). If the MTTDL is acceptable, parity updates are deferred until the array is idle. To that end, the I/O patterns to and from the array are monitored to determine idleness. If the I/O patterns indicate that the disk array is idle, new data is read and parity information is generated which is then written to the array.

Claims

1. A method for storing data in a fault-tolerant storage device having a plurality of disk units forming a disk array with a plurality of stripes for storing data and parity information, comprising the steps of:

identifying at least one target area on said array to write new data, with said at least one target area including old data and being associated with one of said plurality of stripes, defining an updated stripe;

in the absence of a reading step, writing new data to said at least one target area, said new data not having corresponding parity information defining unprotected data;

monitoring a quantity of said unprotected data present on said array to detect presence of a preselected condition related to susceptibility of said at least one target area to loss of said unprotected data; and

in response to said step of monitoring said quantity of said unprotected data and prior to completion of writing all of said new data at said at least one target area, dynamically adjusting said quantity of unprotected data present on said array by generating and writing parity information corresponding to a first portion of said new data to said disk array, while leaving a second portion of said new data as unprotected data.

2. The method as recited in claim 1 wherein said adjusting step adjusts said quantity of unprotected data to ensure said quantity does not exceed a predetermined threshold.

3. The method as recited in claim 1 wherein said monitoring step includes continuously determining a mean-time-to-data-loss.

4. The method as recited in claim 1 wherein said monitoring step includes monitoring said quantity over a predetermined amount of time, to determine an average amount of unprotected data during said period, determining a mean-data-loss-rate.

5. The method as recited in claim 1 wherein said adjusting step includes generating said parity information associated with said new data and writing said parity information to said updated stripe.

6. The method as recited in claim 5 further including the steps of (a) monitoring the rate of I/O requests to the array and (b) generating a parity block associated with said new data upon detection that said rate of I/O requests is within a predetermined range.

7. The method as recited in claim 6 wherein said identifying step and said writing new data step occur a plurality of times, thereby updating a plurality of stripes with new data before said generating said parity block step occurs, with said generating step concurrently producing parity information for said plurality of stripes containing new data.

8. The method as recited in claim 5 further including the step of writing, to memory, information indicating that parity information, corresponding to said new data, is not present on said disk array with said information defining a need-parity-rebuild bit (NPRB), said monitoring step sensing said NPRB and said adjusting step producing said parity information in response to the presence of said NPRB bit.

9. A method for storing data in a fault-tolerant storage device having a plurality of disk drive units forming a disk array with a plurality of stripes for storing data and parity information, comprising the steps of:

monitoring the I/O patterns to and from the disk drive units of the array to determine the idleness of the disk array;

identifying multiple target areas on each disk of said array to write new data, with at least one of said multiple target areas including old data, with said multiple target areas being associated with one of said plurality of stripes, defining a target stripe;

in the absence of a reading step, writing new data to a plurality of said multiple target areas, wherein new data not having corresponding parity information stored on said array defines unprotected data;

monitoring a quantity of unprotected data present on said array to detect when said quantity exceeds a preselected threshold; and

in response to said monitoring of said quantity of unprotected data, dynamically adjusting the quantity of unprotected data by generating and writing to said target stripe parity information corresponding to said unprotected data to at least one of said multiple target areas such that said quantity is maintained at a level below said preselected threshold.

10. The method as recited in claim 9 further including the step of writing to memory, information indicating that parity information, corresponding to said new data, is not present on said array with said information defining a needs parity rebuild bit (NPRB), said generating step producing parity information in response to the presence of said NPRB.

11. The method as recited in claim 9 wherein said generating and writing step generates and writes parity information to ensure said quantity is within a predetermined range.

12. The method as recited in claim 9 wherein said monitoring step includes continuously determining a mean-time-to-data-loss and said generating and writing step generates and writes parity information to ensure said mean-time-to-data-loss is within a predetermined range.

13. The method as recited in claim 9 wherein said monitoring step includes monitoring said quantity over a predetermined amount of time, to determine an average amount of unprotected data during said period, determining a mean-data-loss-rate, and said generating and writing step includes generating and writing parity information to ensure mean-data-loss-rate is within a predetermined range.

14. The method as recited in claim 9 wherein said identifying and writing steps occur a plurality of times, writing a plurality of data stripes to said array before said generating step occurs, with said generating and writing step concurrently producing parity information for said plurality of data stripes and concurrently writing said parity information to said array.

15. An apparatus for generating parity information for a plurality of disk drives, comprising:

a plurality of individual disk drives defining an array of disks having a plurality of stripes, each of said plurality of stripes having multiple target areas in which data may be written, with at least one of said multiple target areas including old data, defining an old data target area;

a controller means, in data communication with said array, for coordinating the operation of said array;

means, in data communication with both said controller means and said array, for writing new data onto a first plurality of said multiple target areas, including said old data target area, defining new data target areas;

means, in data communication with said controller means, for generating parity information from said new data present in said new data target areas, whereby new data not having corresponding parity information written on said array defines unprotected data;

means, in data communication with both said controller means and said array, for monitoring a quantity of unprotected data present on said array to detect presence of a preselected condition relating to susceptibility of said new data target areas to loss of said new data;

means, in data communication with both controller means and said array and responsive to said monitoring means, for dynamically adjusting said quantity of unprotected data on said array by generating and writing parity information corresponding to a first portion of said new data to said array, while leaving a second portion of said new data as unprotected data.

16. The apparatus as recited in claim 15 wherein said adjusting means adjusts said quantity of unprotected data to ensure said quantity is within a predetermined range.

17. The apparatus as recited in claim 15 wherein said monitoring means continuously determines a mean-time-to-data-loss and said adjusting means adjusts said quantity of unprotected data to ensure said mean-time-to-data-loss is within a predetermined range.

18. The apparatus as recited in claim 15 wherein said monitoring means monitors said quantity over a predetermined amount of time, to determine an average amount of unprotected data during said period, and determines a mean-data-loss-rate, and said adjusting means adjusts said quantity to ensure mean-data-loss-rate is within said predetermined range.