Multi level data recovery in storage disk arrays

• US 9,563,524 B2
• Filed: 12/11/2014
• Issued: 02/07/2017
• Est. Priority Date: 12/11/2014
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Goldberg, Itzhack, Martens, Holger, Muehge, Thorsten, Rueger, Erik
• Primary Examiner(s): Lottich, Joshua P
• Application Number:
  US14/567,966
• Publication Number:
  US 20160170851A1
• Time in Patent Office: 789 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 11/0727 in a storage system, e.g. in a DASD or network based storage system drivers for digital recording or reproducing units G06F3/06; circuits for error detection or correction within digital recording or reproducing units G11B20/18; for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS], H04L67/1097
• CPC Class Codes: G06F 11/0754 by exceeding limits
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 11/2069 Management of state, configuration or failover
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/064 Management of blocks
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

Methods, computing systems and computer program products implement embodiments of the present invention that include arranging multiple storage devices as a RAID array and receiving, from a given storage device storing a dataset, a number of error messages exceeding a threshold. A mirrored subset of the dataset is retrieved from the given device, and a calculated subset of the dataset is retrieved from remaining devices in the RAID array. In a first embodiment, the mirrored subset is created first, and the calculated subset includes a relative complement of the mirrored subset with respect to the dataset. In a second embodiment, the calculated subset is created first, and the mirrored subset includes a relative complement of the calculated subset with respect to the dataset. The mirrored and the calculated datasets are stored to an additional storage device, and the given device is replaced with the additional device in the RAID array.

Claims

1. A method, comprising:

arranging multiple storage devices as a Redundant Array of Independent Disks (RAID) array;

receiving, from a given storage device, a number of error messages exceeding a threshold, the given storage device storing a dataset;

retrieving, from the given storage device, a mirrored subset of the dataset;

retrieving, using remaining storage devices in the RAID array, the remaining storage devices comprising the multiple storage devices absent the given storage device, a calculated subset of the dataset, the dataset comprising a union of the mirrored and the calculated subsets; and

creating, on an additional storage device not comprised in the multiple storage devices, a copy of the dataset using the mirrored and the calculated subsets;

wherein each of the storage devices in the RAID array have corresponding storage regions, and wherein retrieving the calculated subset comprises successfully retrieving user data and parity data from a first number of the corresponding storage regions in each of the remaining storage devices in the RAID array, calculating values based on the successfully retrieved user and parity data, storing the calculated values to the calculated subset, and identifying a second number of the corresponding storage regions whose data could not be successfully retrieved; and

wherein retrieving the mirrored subset comprises retrieving, for each of the second number of the corresponding storage regions, data from the corresponding storage region in the given storage device, and storing, to the mirrored subset, the data retrieved from the given storage region.

2. The method according to claim 1, and comprising replacing, in the RAID array, the given storage device with the additional storage device upon creating the copy of the dataset on the additional storage device.

3. The method according to claim 1, wherein each of the multiple and the additional storage devices is selected from a list consisting of a hard disk drive and a solid-state disk drive.

4. The method according to claim 1, wherein each of the storage devices in the RAID array have corresponding storage regions, and wherein retrieving the mirrored subset comprises retrieving data from a first number of the storage regions on the given storage device that are readable, storing the retrieved data to the mirrored subset, and identifying a second number of the storage regions on the given storage device that are unreadable.

5. The method according to claim 4, wherein retrieving the calculated subset comprises retrieving, for each of the second number of the storage regions, user data and parity data from corresponding storage regions on the remaining storage devices in the RAID array, calculating a value based on the user and the parity data retrieved from the corresponding storage regions on the remaining storage devices in the RAID array, and storing the calculated value to the calculated subset.

6. A storage system, comprising:

multiple storage devices arranged as a Redundant Array of Independent Disks (RAID) array;

one or more additional storage devices; and

a processor configured;

to receive, from a given storage device in the RAID array, a number of error messages exceeding a threshold, the given storage device storing a dataset;

to retrieve, from the given storage device in the RAID array, a mirrored subset of the dataset;

to retrieve, using remaining storage devices in the RAID array, the remaining storage devices comprising the multiple storage devices absent the given storage device, a calculated subset of the dataset, the dataset comprising a union of the mirrored and the calculated subsets; and

to create, on a given additional storage device not comprised in the multiple storage devices, a copy of the dataset using the mirrored and the calculated subsets;

wherein each of the storage devices in the RAID array have corresponding storage regions, and wherein the processor is configured to retrieve the calculated subset by successfully retrieving user and parity data from a first number of the corresponding storage regions in each of the remaining storage devices in the RAID array, calculating values based on the successfully retrieved user and parity data, storing the calculated values to the calculated subset, and identifying a second number of the corresponding storage regions whose data could not be successfully retrieved; and

wherein the processor is configured to retrieve the mirrored subset by retrieving, for each of the second number of the corresponding storage regions, data from the corresponding storage region in the given storage device, and storing, to the mirrored subset, the data retrieved from the given storage region.

7. The storage system according to claim 6, wherein the processor is configured to replace, in the RAID array, the given storage device with the additional storage device upon creating the copy of the dataset on the additional storage device.

8. The storage system according to claim 6, wherein each of the multiple and the additional storage devices is selected from a list consisting of a hard disk drive and a solid-state disk drive.

9. The storage system according to claim 6, wherein each of the storage devices in the RAID array have corresponding storage regions, and wherein the processor is configured to retrieve the mirrored subset by retrieving data from a first number of the storage regions on the given storage device that are readable, storing the retrieved data to the mirrored subset, and identifying a second number of the storage regions on the given storage device that are unreadable.

10. The storage system according to claim 9, wherein the processor is configured to retrieve the calculated subset by retrieving, for each of the second number of the storage regions, user data and parity data from corresponding storage regions on the remaining storage devices in the RAID array, calculating a value based on the user and the parity data retrieved from the corresponding storage regions on the remaining storage devices in the RAID array, and storing the calculated value to the calculated sub set.

11. A computer program product, the computer program product comprising:

a non-transitory computer readable storage medium having computer readable program code embodied therewith, the computer readable program code comprising;

computer readable program code configured to arrange multiple storage devices as a Redundant Array of Independent Disks (RAID) array;

computer readable program code configured to receive, from a given storage device, a number of error messages exceeding a threshold, the given storage device storing a dataset;

computer readable program code configured to retrieve, from the given storage device, a mirrored subset of the dataset;

computer readable program code configured to retrieve, using remaining storage devices in the RAID array, the remaining storage devices comprising the multiple storage devices absent the given storage device, a calculated subset of the dataset, the dataset comprising a union of the mirrored and the calculated subsets; and

computer readable program code configured to create, on an additional storage device not comprised in the multiple storage devices, a copy of the dataset using the mirrored and the calculated subsets;

wherein each of the storage devices have corresponding storage regions, and wherein the computer readable program code is configured to retrieve the calculated subset by successfully retrieving user and parity data from a first number of the corresponding storage regions in each of the remaining storage devices in the RAID array, calculating values based on the successfully retrieved user and parity data, storing the calculated values to the calculated subset, and identifying a second number of the corresponding storage regions whose data could not be successfully retrieved; and

wherein the computer readable program code is configured to retrieve the mirrored subset by retrieving, for each of the second number of the corresponding storage regions, data from the corresponding storage region in the given storage device, and storing, to the mirrored subset, the data retrieved from the given storage region.

12. The computer program product according to claim 11, and comprising computer readable program code configured to replace, in the RAID array, the given storage device with the additional storage device upon creating the copy of the dataset on the additional storage device.

13. The computer program product according to claim 11, wherein each of the storage devices in the RAID array have corresponding storage regions, and wherein the computer readable program code is configured to retrieve the mirrored subset by retrieving data from a first number of the storage regions on the given storage device that are readable, storing the retrieved data to the mirrored subset, and identifying a second number of the storage regions on the given storage device that are unreadable.

14. The computer program product according to claim 13, wherein the computer readable program code is configured to retrieve the calculated subset by retrieving, for each of the second number of the storage regions, user data and parity data from corresponding storage regions on the remaining storage devices in the RAID array, calculating a value based on the user and the parity data retrieved from the corresponding storage regions on the remaining storage devices in the RAID array, and storing the calculated value to the calculated subset.