Storage device array architecture with solid-state redundancy unit

• US 6,289,471 B1
• Filed: 12/27/1995
• Issued: 09/11/2001
• Est. Priority Date: 09/27/1991
• Status: Expired due to Term

• Current Assignee: EMC Corporation (Dell Technologies Inc.)
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Gordon, David W.
• Primary Examiner(s): LE, DIEU MINH T
• Application Number:
  US08/579,552
• Time in Patent Office: 2,085 Days
• Field of Search: 395/182.04, 395/182.09, 395/429, 395/431, 395/440, 395/441, 395/444, 395/446, 395/449, 371/40.1, 371/40.4, 371/21.1, 371/22.2, 364/368.3, 364/368.9, 364/242.91, 364/243.4, 364/243.41, 364/244.3, 714/6, 714/11, 714/718, 714/7, 714/8, 714/54, 714/719, 714/723
• US Class Current: 714/6.12
• CPC Class Codes: G06F 11/1008 in individual solid state devices G06F11/1004 takes precedence
• CPC Class Codes: G06F 11/108 Parity data distribution in semiconductor storages, e.g. in SSD
• CPC Class Codes: G06F 11/1435 using file system or storage system metadata

Abstract

A fault tolerant storage subsystem includes a first tier of failure independent data storage units coupled to a storage controller, and a second tier, including at least one failure independent data storage unit, coupled to at least one of the failure independent data storage units in the first tier for providing further fault tolerance thereto. The first tier of failure independent data storage units may include a partially heterogeneous group of data storage units, which may include a solid state data storage unit. Data blocks representative of data received from the storage controller are stored on at least a first subset of the first tier of failure independent data storage units and associated redundancy blocks representative of redundancy information associates with the data blocks are stored on a second subset of the first tier of failure independent data storage units.

Claims

1. A fault tolerant storage subsystem comprising:

a first tier of failure independent data storage units coupled to a storage controller; and

a second tier of at least one failure independent data storage units coupled to at least one of said failure independent data storage units in said first tier for providing further fault tolerance thereto.

2. The fault tolerant storage subsystem of claim 1 wherein said first tier of failure independent data storage units comprises an at least partially heterogeneous group of data storage units.

3. The fault tolerant storage subsystem of claim 2 wherein at least one of said first tier of failure independent data storage units comprises a solid state data storage unit.

4. The fault tolerant storage subsystem of claim 3 wherein said at least one failure independent data storage unit in said second tier is coupled to said solid state data storage unit.

5. The fault tolerant storage subsystem of claim 4 wherein said at least one failure independent data storage unit in said second tier comprises a solid state data storage unit.

6. The fault tolerant storage subsystem of claim 1 wherein said data blocks representative of data received from said storage controller are stored on at least a first subset of said first tier of failure independent data storage units and associated redundancy blocks representative of redundancy information associated with said data blocks are stored on a second subset of said first tier of failure independent data storage units.

7. The fault tolerant storage subsystem of claim 6 further comprising:

at least one redundancy block calculation circuit coupled to said first subset of said first tier of failure independent data storage units for calculating said associated redundancy blocks for storage on said second subset of said first tier of failure independent data storage units.

8. The fault tolerant storage subsystem of claim 6 wherein said associated redundancy blocks stored on said second subset of said first tier of failure independent data storage units are computed based upon said data blocks in said first subset of said first tier of failure independent data storage units.

9. The fault tolerant storage subsystem of claim 6 wherein said second tier of at least one failure independent data storage units coupled to at least one of said failure independent data storage units in said first tier comprises a mirrored storage unit.

10. The fault tolerant storage subsystem of claim 6 further comprising a buffer memory associated with at least said first subset of said first tier of failure independent data storage units for temporary storage of said data stored on said first subset of said first tier of failure independent data storage units.

11. The fault tolerant storage subsystem of claim 6 further comprising a buffer memory associated with at least said second subset of said first tier of failure independent data storage units for temporary storage of said data stored on said second subset of said first tier of failure independent data storage units.

12. A storage subsystem comprising:

a storage controller coupling a central processing unit to a plurality of failure independent data storage units, at least one of said plurality of failure independent data storage units comprising a solid state data storage unit; and

a non-volatile solid state buffer memory associated with at least said solid state data storage unit to provide a data cache;

wherein said plurality of failure independent data storage units comprises a first tier of failure independent data storage units coupled to said storage controller and a second tier of at least one failure independent data storage units coupled to at least one of said failure independent data storage units in said first tier for providing further fault tolerance thereto.

13. The storage subsystem of claim 12 wherein said first tier of failure independent data storage units comprises an at least partially heterogeneous group of data storage units.

14. The storage subsystem of claim 12 wherein said at least one failure independent data storage unit in said second tier is coupled to said solid state data storage unit.

15. The storage subsystem of claim 14 wherein said at least one failure independent data storage unit in said second tier comprises a solid state data storage unit.

16. The storage subsystem of claim 12 wherein data blocks representative of data received from said storage controller are stored on at least a first subset of said failure independent data storage units and associated redundancy blocks representative of redundancy information associated with said data blocks are stored on a second subset of said failure independent data storage unit.

17. The storage subsystem of claim 16 further comprising:

at least one redundancy block calculation circuit coupled to said first subset of failure independent data storage units for calculating said associated redundancy blocks for storage on said second subset of said failure independent data storage units.

18. The storage subsystem of claim 17 wherein said associated redundancy blocks stored on said second subset of said failure independent data storage units are computed based upon said data blocks in said first subset of said failure independent data storage units.

19. The storage subsystem of claim 12 wherein said second tier of at least one failure independent data storage units coupled to at least one of said failure independent data storage units in said first tier comprises a mirrored storage unit.

20. The subsystem of claim 16 further comprising a buffer memory associated with at least said first subset of failure independent data storage units for temporary storage of said data stored on said first subset of failure independent data storage units.

21. The storage subsystem of claim 16 further comprising a buffer memory associated with at least said second subset of failure independent data storage units for temporary storage of said data stored on said second subset of failure independent data storage units.

22. A storage subsystem comprising:

a storage controller coupling a central processing unit to a plurality of data storage units in a redundant array, at least one of said data storage units having an access time that is different from at least another of said storage units, said storage controller operative for placing primarily accessed data on a fastest one of said failure independent data storage units.

23. The storage subsystem of claim 22 wherein said fastest one of said failure independent data storage units comprises a solid state data storage unit.

24. The storage subsystem of claim 22 further comprising a separate tier of at least one failure independent data storage unit coupled to at least one of said plurality of at least partially differentially Read and Write cycle time accessible failure independent data storage units.

25. The storage subsystem of claim 24 wherein said at least one failure independent data storage unit in said separate tier comprises a solid state data storage unit.

26. The storage subsystem of claim 24 wherein said at least one failure independent data storage unit in said separate tier is coupled to a solid state data storage unit.

27. The storage subsystem of claim 24 wherein said separate tier of at least one failure independent data storage units coupled to at least one said plurality of failure independent data storage units comprises a mirrored storage unit.

28. The storage subsystem of claim 22 wherein data blocks representative of data received from said storage controller are stored on at least a first subset of said failure independent data storage units and associated redundancy blocks representative of redundancy information associated with said data blocks are stored on a second subset of said failure independent data storage units.

29. The storage subsystem of claim 28 further comprising:

at least one redundancy block calculation circuit coupled to said first subset of failure independent data storage units for calculating said associated redundancy blocks for storage on said second subset of failure independent data storage units.

30. The storage subsystem of claim 28 wherein said associated redundancy blocks stored on said second subset of failure independent data storage units are computed based upon said data blocks in said first subset of failure independent data storage units.

31. The storage subsystem of claim 28 further comprising a buffer memory associated with at least said first subset of said failure independent data storage units for temporary storage of said data stored on said first subset of failure independent data storage units.

32. The storage subsystem of claim 28 further comprising a buffer memory associated with at least said second subset of failure independent data storage units for temporary storage of said data stored on said second subset of failure independent data storage units.

33. A fault tolerant storage subsystem comprising:

a plurality of failure independent data storage units for storing a plurality of data blocks and at least one associated redundancy block, at least one of said plurality of failure independent data storage units comprising a solid state data storage unit having a Read and Write cycle time substantially lesser than a Read and Write cycle time of each other of said failure independent data storage units and wherein selected ones of said plurality of said data blocks and said at least one associated redundancy block which are most frequently accessed during a data modification operation are stored on said at least one solid state data storage unit to decrease Read and Write operation times of said frequently accessed data to that of said solid state data storage unit.

34. The fault tolerant storage subsystem of claim 33 wherein said most frequently accessed one of said plurality of said data blocks and said at least one associated redundancy block is said at least one associated redundancy block.

35. The fault tolerant storage subsystem of claim 33 wherein said plurality of failure independent data storage units are arranged in a first tier coupled to a storage controller.

36. The fault tolerant storage subsystem of claim 35 further comprising a second tier of at least one failure independent data storage units coupled to at least one of said failure independent data storage units in said first tier.

37. The fault tolerant storage subsystem of claim 36 wherein said at least one failure independent data storage unit of said second tier is coupled to said solid state data storage unit.

38. The fault tolerant storage subsystem of claim 37 wherein said at least one failure independent data storage unit of said second tier comprises a mirrored storage unit.

39. The fault tolerant storage subsystem of claim 37 further comprising:

at least one redundancy block calculation circuit coupled to said first subset of failure independent data storage units for calculating said associated redundancy blocks for storage on said second subset of said failure independent data storage units.

40. The fault tolerant storage subsystem of claim 39 wherein said associated redundancy blocks stored on said second subset of said failure independent data storage units are computed based upon said data blocks in said first subset of said failure independent data storage units.

41. The fault tolerant storage subsystem of claim 33 wherein data blocks representative of data received from a storage controller are stored on at least a first subset of said failure independent data storage units and associated redundancy blocks representative of redundancy information associated with said data blocks are stored on a second subset of said failure independent data storage units.

42. The fault tolerant storage subsystem of claim 41 further comprising a buffer memory associated with at least said first subset of said failure independent data storage units for temporary storage of said data stored on said first subset of failure independent data storage units.

43. The fault tolerant storage subsystem of claim 41 further comprising a buffer memory associated with at least said second subset of failure independent data storage units for temporary storage of said data stored on said second subset of failure independent data storage units.

44. A computer comprising:

a central processing unit;

a storage subsystem controller coupled to said central processing unit;

a plurality of first tier failure independent data storage units coupled to said controller in a redundant array, at least one of said plurality of failure independent data storage units having a Read and Write cycle time substantially differing from at least one other of said failure independent data storage units.

45. The computer system of claim 44 wherein said at least one of said plurality of failure independent data storage units comprises a solid state data storage unit.

46. The computer system of claim 44 wherein said plurality of failure independent data storage units further comprise at least one additional data storage unit coupled to one of said failure independent data storage units to provide fault tolerance thereto.

47. The computer system of claim 46 wherein said at least one additional data storage unit is coupled to one of said failure independent data storage units through an additional storage subsystem controller.

48. The computer system of claim 46 wherein said at least one additional data storage unit comprises a solid state data storage unit.

49. The computer system of claim 44 wherein data blocks representative of data received from said storage subsystem controller are stored on at least a first subset of said plurality of failure independent data storage units and associated redundancy blocks representative of redundancy information associated with said data blocks are stored on a second subset of said plurality of failure independent data storage units.

50. The computer system of claim 49 further comprising:

at least one redundancy block calculation circuit coupled to said first subset of said plurality of failure independent data storage units for calculating said associated redundancy blocks for storage on said second subset of said plurality of failure independent data storage units.

51. The computer system of claim 50 wherein said associated redundancy blocks stored on said second subset of said plurality of failure independent data storage units are computed based upon said data blocks in said first subset of said plurality of failure independent data storage units.

52. A fault tolerant storage subsystem comprising:

a plurality of failure independent data storage units coupled to a storage controller in a redundant array, said plurality of failure independent data storage units comprising at least first and second data storage units having substantially differing Read and Write cycle times.

53. The fault tolerant storage subsystems of claim 52 wherein said first data storage unit comprises a solid state data storage unit and said second data storage unit comprises a rotating randomly accessible storage medium.

54. The fault tolerant storage subsystem of claim 52 wherein said plurality of failure independent data storage units further comprise at least one additional data storage unit coupled to one of said plurality of failure independent data storage units to provide fault tolerance thereto.

55. The fault tolerant storage subsystem of claim 54 wherein said at least one additional data storage unit is coupled to one of said plurality of failure independent data storage units through an additional storage controller.

56. The fault tolerant storage subsystem of claim 52 wherein data blocks representative of data received from said storage controller are stored on at least a first subset of said plurality of failure independent data storage units and associated redundancy blocks representative of redundancy information associated with said data blocks are stored on a second subset of said plurality of failure independent data storage units.

57. The fault tolerant storage subsystem of claim 56 further comprising:

at least one redundancy block calculation circuit coupled to said first subset of said plurality of failure independent data storage units for calculating said associated redundancy blocks for storage on said second subset of said plurality of failure independent data storage units.

58. The fault tolerant storage subsystem of claim 57 wherein said associated redundancy blocks stored on said second subset of said plurality of failure independent data storage units are computed based upon said data blocks in said first subset of said plurality of failure independent data storage units.

59. A method for increasing performance of a fault tolerant storage subsystem comprising the steps of:

providing a plurality of failure independent data storage units for storing a plurality of data blocks and at least one associated redundancy block, at least one of said plurality of failure independent data storage units comprising at least one data storage unit having a Read and Write cycle time substantially lesser than a Read and Write cycle time of at least one other of said failure independent data storage units;

determining which of said plurality of data blocks and said at least one associated redundancy block are most frequently accessed during a data modification operation; and

storing said most frequently accessed ones of said plurality of data blocks and said at least one associated redundancy block on said at least one data storage unit.

60. The method of claim 59 wherein said step of providing is carried out by at least one solid state data storage unit.

61. The method of claim 60 further comprising the step of:

coupling a buffer memory between at least said solid state data storage unit and a storage controller.

62. The method of claim 59 further comprising the step of:

additionally providing at least one failure independent data storage unit coupled to at least one of said plurality of failure independent data storage units for providing further fault tolerance thereto.

63. The method of claim 62 further comprising the step of:

additionally providing a second storage subsystem to couple at least one additional failure independent storage unit to at least one of said plurality of failure independent data storage units for providing further fault tolerance thereto.

64. The method of claim 62 wherein said step of additionally providing is carried out by at least one solid state data storage unit.

65. The method of claim 62 wherein said step of additionally providing is carried out by at least one pair of mirrored failure independent data storage units.