Distributing workload evenly across storage media in a storage array

• US 20030212872A1
• Filed: 05/08/2002
• Published: 11/13/2003
• Est. Priority Date: 05/08/2002
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Patterson, Brian, Condel, Jonathan
• Granted Patent: US 6,912,635 B2
• Time in Patent Office: Days
• US Class Current: 711/165
• CPC Class Codes: G06F 2003/0697 device management, e.g. handlers, drivers, I/O schedulers
• CPC Class Codes: G06F 3/0601 Interfaces specially adapted for storage systems
• CPC Class Codes: G06F 3/0613 in relation to throughput
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/0653 Monitoring storage devices or systems
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

A system and methods employ a redistribution module to predict whether there is data stored in an arrayed storage device that is likely to be data that will be highly accessed in the future. The redistribution module locates high-access data on one or more storage components within the array and redistributes it evenly across all storage components in the array so that no single storage component contains a disproportionate amount of the high-access data. Redistributing high-access data evenly across all storage components (e.g., disks) in a storage array helps prevent I/O (input/output) bottlenecking at any single storage component in the array. I/Os are serviced in a more parallel manner which increases the overall performance of the storage array.

Claims

1. A processor-readable medium comprising processor-executable instructions configured for:

compiling workload information for storage components within a storage array;

based on the workload information, predicting that at least one storage component contains high-access data; and

redistributing the high-access data evenly across all storage components within the storage array.

2. A processor-readable medium as recited in claim 1, wherein the redistributing further comprises:

locating the high-access data on the at least one storage component;

removing high-access data from the at least one storage component; and

storing high-access data evenly onto all storage components within the storage array.

3. A processor-readable medium as recited in claim 2, wherein the removing high-access data comprises removing all of the high-access data.

4. A processor-readable medium as recited in claim 2, wherein the removing high-access data comprises removing some of the high-access data.

5. A processor-readable medium as recited in claim 1, wherein the predicting further comprises:

accessing the workload information; and

executing a predictive algorithm that operates on the workload information.

6. A processor-readable medium as recited in claim 1, wherein the redistributing further comprises redistributing during a least disruptive time.

7. A processor-readable medium as recited in claim 6, comprising further processor-executable instructions configured for:

monitoring a utilization rate of the storage array; and

based on the utilization rate, determining the least disruptive time.

8. A processor-readable medium as recited in claim 1, wherein the compiling further comprises storing the workload information in a memory on the storage array.

9. A processor-readable medium as recited in claim 1, wherein the storage components are selected from a group of components comprising:

magnetic disks;

tapes;

optical disks; and

solid state disks.

10. A processor-readable medium comprising processor-executable instructions configured for:

compiling workload information for storage components within a storage array;

based on the workload information, determining least busy storage components and most busy storage components;

determining that a new disk access is a random I/O (input/output); and

using a least busy storage component to service the new disk access.

11. A processor-readable medium in a storage array having a plurality of storage components, the processor-readable medium comprising processor-executable instructions configured for:

accessing foreknowledge information to determine that data on at least one storage component is high-access data; and

redistributing the high-access data evenly across all storage components within the storage array.

12. A processor-readable medium as recited in claim 11, wherein the redistributing further comprises:

locating the high-access data on the at least one storage component;

removing high-access data from the at least one storage component; and

storing high-access data evenly onto all storage components within the storage array.

13. A processor-readable medium as recited in claim 12, wherein the removing high-access data comprises removing all of the high-access data.

14. A processor-readable medium as recited in claim 12, wherein the removing high-access data comprises removing some of the high-access data.

15. A processor-readable medium as recited in claim 11, comprising further processor-executable instructions configured for:

prior to the accessing, storing the foreknowledge information in a memory on the storage array; and

wherein the accessing comprises accessing the foreknowledge information from the memory.

16. A processor-readable medium as recited in claim 11, wherein the redistributing further comprises redistributing during a least disruptive time.

17. A processor-readable medium as recited in claim 16, comprising further processor-executable instructions configured for:

monitoring a utilization rate of the storage array; and

based on the utilization rate, determining the least disruptive time.

18. A processor-readable medium as recited in claim 11, wherein the storage components are selected from a group of components comprising:

magnetic disks;

tapes;

optical disks; and

solid state disks.

19. A processor-readable medium in an array of disks, the processor-readable medium comprising processor-executable instructions configured for:

predicting that at least one disk contains high-access data; and

redistributing the high-access data evenly across all disks within the array.

20. A processor-readable medium as recited in claim 19, wherein the predicting further comprises:

compiling workload information that indicates the workload for each disk within the array; and

executing a predictive algorithm that operates on the workload information.

21. A processor-readable medium as recited in claim 19, wherein the predicting further comprises accessing previously stored foreknowledge information that indicates data that is high-access data.

22. A processor-readable medium as recited in claim 19, wherein the redistributing further comprises:

locating the high-access data on the at least one disk;

removing high-access data from the at least one disk; and

storing high-access data evenly onto all disks within the array of disks.

23. A processor-readable medium as recited in claim 19, wherein the redistributing further comprises redistributing during a least disruptive time.

24. A processor-readable medium as recited in claim 23, comprising further processor-executable instructions configured for:

monitoring a utilization rate of the array of disks; and

based on the utilization rate, determining the least disruptive time.

25. A method of distributing workload evenly across storage components in a storage array comprising:

compiling workload information for each storage component within the storage array;

based on the workload information, predicting that at least one storage component contains high-access data; and

redistributing the high-access data evenly across all storage components within the storage array.

26. A method of distributing workload evenly across storage components in a storage array comprising:

compiling workload information for each storage component within the storage array;

based on the workload information, determining least busy storage components and most busy storage components;

determining that a new disk access is a random I/O (input/output); and

using a least busy storage component to service the new disk access.

27. A method of distributing workload evenly across storage components in a storage array comprising:

accessing foreknowledge information to determine that data on at least one storage component is high-access data; and

redistributing the high-access data evenly across all storage components within the storage array.

28. A method of distributing workload evenly across disks in an array of disks comprising:

predicting that at least one disk contains high-access data; and

redistributing the high-access data evenly across all disks within the array.

29. A storage array comprising:

two or more storage components;

a redistribution module configured to access storage component workload information, and based on the workload information, to redistribute high-access data evenly across the two or more storage components.

30. A storage array as recited in claim 29, further comprising a predictive algorithm configured as part of the redistribution module to analyze the workload information and predict the high-access data.

31. A storage array as recited in claim 29, wherein the two or more storage components are selected from a group of storage components comprising:

magnetic disks;

tapes;

optical disks; and

solid state disks.

32. A storage array comprising:

a processor;

two or more storage components;

a memory having foreknowledge information stored therein; and

a redistribution module executable on the processor to access the memory, determine from the foreknowledge information which data on the two or more storage components is high-access data, and redistribute the high-access data evenly across the two or more storage components.

33. A redundant array of independent disks comprising:

two or more disks;

a redistribution module configured to determine high-access data stored on a disk and to redistribute the high-access data evenly across the two or more disks.

34. A system comprising:

a host computer;

a virtual storage array to service data requests from the host computer, the virtual storage array having two or more storage components; and

a redistribution module executable on the virtual storage array to access workload information, and based on the workload information, to redistribute high-access data evenly across the two or more storage components.