Disk array having hot spare resources and methods for using hot spare resources to store user data

• US 5,666,512 A
• Filed: 02/10/1995
• Issued: 09/09/1997
• Est. Priority Date: 02/10/1995
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.), Dystar Textilfarben GmbH & Company Deutschland KG (Zhejiang Longsheng Group Company Limited)
• Original Assignee: Hewlett-Packard Company (HP Inc.)
• Inventors: Burkes, Theresa A., Voigt, Douglas L., Jacobson, Michael B., Nelson, Marvin D., Dolphin, Wade A., Diamond, Bryan M.
• Primary Examiner(s): Swann, Tod R.
• Assistant Examiner(s): SAUNDERS, KEITH
• Application Number:
  US08/386,574
• Time in Patent Office: 942 Days
• Field of Search: 395/441, 395/488, 395/489, 395/497.04, 395/182.03, 395/182.04, 395/182.05
• US Class Current: 711/114
• 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 2211/1028 Distributed, i.e. distributed RAID systems with parity
• CPC Class Codes: G06F 2211/1059 Parity-single bit-RAID5, i.e. RAID 5 implementations
• CPC Class Codes: G11C 29/74 using duplex memories, i.e. using dual copies

Abstract

A disk array data storage system has a plurality of storage disks and a disk array controller for coordinating transfer of user data to and from the storage disks. A memory manager is provided to manage memory allocation and data storage on the disk array. The memory manager maintains a sufficient quantity of hot spare storage space that can be made available for reconstructing the user data and restoring redundancy in the event one of the storage disks fails. Until a disk fails, however, the memory manager uses the hot spare space to store user and redundant data while guaranteeing that the storage space can be used for rebuilding following a storage disk failure. In this manner, all storage disks in the array are used to store user data, and additionally to maintain storage space that can be easily made available for purposes of hot spare in the event of disk failure.

Claims

1. A method for providing storage space in a hierarchic storage disk array for use in rebuilding in the event a storage disk fails, the disk array having a plurality of storage disks that define a physical storage space the physical storage space being mapped into a RAID-level virtual storage space that presents the physical storage space as multiple RAID areas including mirror and parity RAID areas, the method comprising the following steps:

prior to a disk failure, storing user data in both the mirror and parity RAID areas in a manner that the user data is distributed across all of the storage disks;

storing redundant data on at least one of the storage disks;

ensuring that a sufficient quantity of storage space can be made available for reconstructing the user data and restoring redundancy in the event one of the storage disks fails so that the user data is stored in both the mirror and parity RAID areas on all storage disks before and after the storage disk failure; and

using the quantity of storage space to store both user and redundant data until the storage disk failure occurs.

2. A method according to claim 1 further comprising the additional step of distributing any available storage space that can be used for reconstructing user data and restoring redundancy across all of the storage disks in the disk array.

3. A method according to claim 1 further comprising the step of converting the storage space used to store user and redundant data to hot spare space for reconstructing the user data and restoring redundancy.

4. A method according to claim 1 further comprising the step of generating a warning to inform a user that hot spare resources cannot be guaranteed when it is determined that a sufficient quantity of storage space cannot be made available for reconstructing user data and restoring redundancy in the disk array following a disk failure.

5. A method for providing storage space for use in rebuilding in a hierarchic data storage system, the data storage system comprising a disk array of more than two storage disks of selected capacities that define a physical storage space, the method comprising the following steps:

mapping the physical storage space into a RAID-level virtual storage space which presents the physical storage space as multiple RAID areas, the RAID areas including mirror RAID areas which store data according to mirror redundancy and parity RAID areas which store data according to parity redundancy, the mirror and parity RAID areas extending across all of the storage disks in the disk array;

storing user data in both the mirror and parity RAID areas in a manner that the user data is distributed across all of the storage disks in the disk array;

ensuring that a sufficient quantity of the RAID-level virtual storage space can be made available for reconstructing user data and restoring redundancy in the event of a storage disk failure;

using the quantity of the RAID-level virtual storage space to store data until the storage disk failure occurs; and

in the event of a storage disk failure, performing one of the following two steps;

(a) using unused storage space which is not allocated as mirror or parity RAID areas as the storage space to reconstruct user data and restore redundancy from the failed storage disk, or (b) migrating user data from mirror RAID areas to parity RAID areas and packing the mirror RAID areas to free up storage space to compensate for the space lost from the failed storage disk;

wherein each storage disk stores data according to both mirror and parity redundancy before and after storage disk failure.

6. A method according to claim 5 further comprising the additional step of using the quantity of storage space for reconstructing the user data and restoring redundancy.

7. A method according to claim 5 further comprising the following steps:

mapping the RAID-level virtual storage space into an application-level virtual storage space;

dynamically migrating the user data between the mirror and parity RAID areas; and

computing a capacity of the application-level virtual storage space that can be committed to a user without including a capacity of the largest size storage disk.

8. A method according to claim 7 further comprising the additional step of preventing commitment of additional application-level virtual storage space beyond the computed capacity to ensure that the sufficient quantity of the RAID-level virtual storage space can be made available for rebuilding.

9. A method according to claim 5 further comprising the following step:

in the event that the sufficient quantity of the RAID-level virtual storage space cannot be ensured, generating a warning to inform a user that hot spare resources cannot be guaranteed.

10. A method according to claim 5 wherein any available storage space that can be used for reconstructing user data and restoring redundancy is distributed across multiple ones of the storage disks in the disk array.

11. A data storage system comprising:

a hierarchic disk array having a plurality of storage disks that define a physical storage space;

a disk array controller coupled to the disk array for coordinating transfer of user data to and from the storage disks;

a memory manager operatively coupled to the disk array controller to manage a map of the physical storage space into a RAID-level virtual storage space which presents the physical storage space as multiple RAID areas including mirror RAID areas which store data according to mirror redundancy and parity RAID areas which store data according to parity redundancy, the memory manager storing user data in both the mirror and parity RAID areas in a manner that the user data is distributed across all of the storage disks in the disk array prior to any disk failure; and

the memory manager maintaining a sufficient quantity of storage space that can be made available for reconstructing the user data and restoring redundancy in the event one of the storage disks fails, the memory manager using the quantity of storage space to store both user and redundant data while guaranteeing that the storage space can be made available for rebuilding following a storage disk failure so that each storage disk stores the user data according to both mirror and parity redundancy before and after storage disk failure.

12. A data storage system according to claim 11 wherein the memory manager distributes any available storage space that can be used for reconstructing the user data and restoring redundancy among the storage disks in the disk array.

13. A data storage system according to claim 11 wherein:

the RAID management system stores user data in mirror RAID areas and in the event of a storage disk failure, migrates user data from mirror RAID areas to parity RAID areas to free up storage space to compensate for the space lost from the failed storage disk.

14. A data storage system according to claim 11 wherein:

the RAID management system further mapping the RAID-level virtual storage space into an application-level virtual storage space; and

the RAID management system computing a capacity of the application-level virtual storage space that can be committed to a user without including a capacity of a largest size storage disk.

15. A data storage system according to claim 14 wherein the RAID management system prevents commitment of additional application-level virtual storage space beyond the computed capacity to ensure that the sufficient quantity of the RAID-level virtual storage space can be made available for reconstructing the user data and restoring redundancy.

16. A data storage system according to claim 11 wherein, in the event that the sufficient quantity of the RAID-level virtual storage space cannot be maintained, the RAID management system generates a warning to inform a user that hot spare resources cannot be guaranteed.

17. A method for operating a hierarchic data storage system, the data storage system comprising a disk array of plural storage disks of selected capacities that define a physical storage space, the physical storage space being mapped into a RAID-level virtual storage space which presents the physical storage space as multiple RAID areas, the RAID areas including mirror RAID areas which store data according to mirror redundancy and parity RAID areas which store data according to parity redundancy, the RAID areas extending across the storage disks in the disk arrays, the method comprising the following steps:

storing user data in the minor and parity RAID areas in a manner that the user data is distributed across all of the storage disks in the disk array and all of the storage disks store the user data according to both minor and parity RAID redundancy;

ensuring that a sufficient quantity of the RAID-level virtual storage space can be made available for reconstructing user data and restoring redundancy in the event of a storage disk failure;

dynamically migrating the user data between the mirror and parity RAID areas;

in the event of a storage disk failure, continuously migrating the user data from mirror RAID areas to parity RAID areas and rebuilding redundancy; and

wherein each storage disk stores the user data according to both mirror and parity redundancy before and after storage disk failure.

18. A method according to claim 17 wherein each storage disk stores data according to both RAID 1 and RAID 5 before and after the storage disk failure.

19. A method according to claim 17 further comprising the step of dynamically migrating data between the mirror and parity RAID areas after redundancy is restored following the storage disk failure.

20. A method according to claim 17 further comprising the following steps:

initially storing the user data in mirror RAID areas; and

storing the user data in both mirror and parity RAID areas as the amount of user data exceeds a threshold amount.