APPARATUS AND METHOD TO ALLOCATE RESOURCES IN A DATA STORAGE LIBRARY

• US 20090210618A1
• Filed: 02/14/2008
• Published: 08/20/2009
• Est. Priority Date: 02/14/2008
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: BATES, ALLEN KEITH, WINARSKI, DANIEL JAMES, KLEIN, CRAIG ANTHONY, HAUSTEIN, NILS
• Granted Patent: US 7,979,635 B2
• Time in Patent Office: Days
• US Class Current: 711/114
• CPC Class Codes: G06F 11/1084 Degraded mode, e.g. caused by single or multiple storage removals or disk failures
• CPC Class Codes: G06F 11/1092 Rebuilding, e.g. when physically replacing a failing disk

Abstract

A method to allocate resources in a data storage library comprising a plurality of data storage devices configured as a RAID array, by establishing a normal operation resource allocation, a RAID failure resource allocation, and a multiple storage device failure resource allocation. The method receives host I/O requests, and enqueues those host I/O requests. If the system is operating without any storage device failures, then the method processes host I/O requests using the normal operation resource allocation. If the system is operating with a storage device failure, then the method processes host I/O requests using the RAID failure resource allocation. If the system is operating with multiple storage device failures, then the method processes host I/O requests using the multiple storage device failure resource allocation.

Claims

1. A method to allocate resources in a data storage library comprising a plurality of data storage devices configured as a RAID array, comprising the steps of:

establishing a normal operation resource allocation;

establishing a RAID failure resource allocation;

establishing a multiple storage device failure resource allocation;

receiving host I/O requests;

enqueuing said host I/O requests.

2. The method of claim 1, wherein:

said data storage library comprises a RAID controller in communication with a host computer and interconnected to said plurality of data storage devices;

said RAID controller is capable of performing a nominal number of instructions per second;

said method of claim comprising the steps of;

selecting an enqueued host I/O request;

determining if said RAID array is operating with a RAID failure;

operative if said RAID array is not operating with a RAID failure, processing said selected host I/O request using said normal operation resource allocation.

3. The method of claim 2, wherein said normal operation resource allocation allocates between about 85 percent to about 95 percent of said nominal number of instructions per second to processing host I/O requests.

4. The method of claim 2, further comprising the steps of:

operative if said RAID array is operating with a RAID failure, determining if said RAID array is operating with multiple storage device failures;

operative if said RAID array is operating with a RAID failure but not with multiple storage device failures, processing said elected host I/O request using said RAID failure operation resource allocation.

5. The method of claim 4, wherein said RAID failure operation resource allocation allocates between about 40 percent to about 60 percent of said nominal number of instructions per second to processing host I/O requests.

6. The method of claim 4, further comprising the step of processing said elected host I/O request using said multiple storage device failure operation resource allocation if said RAID array is operating with multiple storage device failures.

7. The method of claim 6, wherein said multiple storage device failure operation resource allocation allocates between about 5 to about 15 percent of said nominal number of instructions per second to processing host I/O requests.

8. The method of claim 4, further comprising the steps of:

establishing a minimum RAID rebuild rate threshold;

establishing a maximum RAID rebuild threshold;

determining an actual RAID rebuild rate;

determining if said actual RAID rebuild rate is greater than said minimum RAID rebuild rate threshold;

operative if said actual RAID rebuild rate is not greater than said minimum RAID rebuild rate threshold, adjusting said RAID failure resource allocation to decrease the system resources allocated for processing of host I/O requests.

9. The method of claim 8, further comprising the steps of:

operative if said actual RAID rebuild rate is greater than said minimum RAID rebuild rate threshold, determining if said actual RAID rebuild rate is less than said maximum RAID rebuild rate;

operative if said actual RAID rebuild rate is not less than said maximum RAID rebuild rate threshold, adjusting said RAID failure resource allocation to decrease the system resources allocated for a RAID rebuild.

10. A RAID controller comprising a processor capable of a nominal number of instructions per second and a computer readable medium comprising a normal operation resource allocation, a RAID failure resource allocation, a multiple storage device failure resource allocation, and computer readable program code disposed therein to allocate a percentage of said nominal number of instructions per second to processing host I/O requests, wherein said RAID controller is in communication with a host computer and interconnected with a plurality of data storage media configured as a RAID array, the computer readable program code comprising a series of computer readable program steps to effect:

receiving host I/O requests;

enqueuing said host I/O requests.

selecting an enqueued host I/O request;

determining if said RAID array is operating with a RAID failure;

operative if said RAID array is not operating with a RAID failure, processing said selected host I/O request using said normal operation resource allocation.

11. The RAID controller of claim 10, said computer readable program code further comprising a series of computer readable program steps to effect:

operative if said RAID array is operating with a RAID failure, determining if said RAID array is operating with multiple storage device failures;

operative if said RAID array is operating with a RAID failure but not with multiple storage device failures, processing said selected host I/O request using said RAID failure operation resource allocation.

12. The RAID controller of claim 11, said computer readable program code further comprising a series of computer readable program steps to effect processing said selected host I/O request using said multiple storage device failure operation resource allocation if said RAID array is operating with multiple storage device failures.

13. The RAID controller of claim 11, wherein said computer readable memory further comprises an encoded minimum RAID rebuild rate threshold and an encoded maximum RAID rebuild threshold, said computer readable program code further comprising a series of computer readable program steps to effect:

determining an actual RAID rebuild rate;

determining if said actual RAID rebuild rate is greater than said minimum RAID rebuild rate threshold;

operative if said actual RAID rebuild rate is not greater than said minimum RAID rebuild rate threshold, adjusting said RAID failure resource allocation to decrease the system resources allocated for processing host I/O requests.

14. The RAID controller of claim 13, said computer readable program code further comprising a series of computer readable program steps to effect:

operative if said actual RAID rebuild rate is greater than said minimum RAID rebuild rate threshold, determining if said actual RAID rebuild rate is less than said maximum RAID rebuild rate;

operative if said actual RAID rebuild rate is not less than said maximum RAID rebuild rate threshold, adjusting said RAID failure resource allocation to decrease the system resources allocated for a RAID rebuild.

15. A computer program product encoded in a computer readable medium, said computer program product being useable by a RAID controller comprising a processor capable of a nominal number of instructions per second and a computer readable medium comprising a normal operation resource allocation, a RAID failure resource allocation, a multiple storage device failure resource allocation, wherein said RAID controller is in communication with a host computer and interconnected with a plurality of data storage media configured as a RAID array, comprising:

computer readable program code which causes said programmable computer processor to receive host I/O requests;

computer readable program code which causes said programmable computer processor to enqueue said host I/O requests.

computer readable program code which causes said programmable computer processor to select an enqueued host I/O request;

computer readable program code which causes said programmable computer processor to determine if said RAID array is operating with a RAID failure;

computer readable program code which, if said RAID array is not operating with a RAID failure, causes said programmable computer processor to process said selected host I/O request using said normal operation resource allocation.

16. The method of claim 15, further comprising:

computer readable program code which, if said RAID array is operating with a RAID failure, causes said programmable computer processor to determine if said RAID array is operating with multiple storage device failures;

computer readable program code which, if said RAID array is operating with a RAID failure but not with multiple storage device failures, causes said programmable computer processor to process said selected host I/O request using said RAID failure operation resource allocation.

17. The method of claim 16, further comprising computer readable program code which causes said programmable computer processor to process said selected host I/O request using said multiple storage device failure operation resource allocation if said RAID array is operating with multiple storage device failures.

18. The method of claim 16, wherein said computer readable memory further comprises a minimum RAID rebuild rate threshold and a maximum RAID rebuild threshold, further comprising:

computer readable program code which causes said programmable computer processor to determine an actual RAID rebuild rate;

computer readable program code which causes said programmable computer processor to determine if said actual RAID rebuild rate is greater than said minimum RAID rebuild rate threshold;

computer readable program code which, if said actual RAID rebuild rate is not greater than said minimum RAID rebuild rate threshold, causes said programmable computer processor to adjust said RAID failure resource allocation to decrease the system resources allocated for processing host I/O requests.

19. The method of claim 18, further comprising

computer readable program code which, if said actual RAID rebuild rate is greater than said minimum RAID rebuild rate threshold, causes said programmable computer processor to determine if said actual RAID rebuild rate is less than said maximum RAID rebuild rate;

computer readable program code which, if said actual RAID rebuild rate is not less than said maximum RAID rebuild rate threshold, causes said programmable computer processor to decrease the system resources allocated for a RAID rebuild.

20. A method to allocate resources in a first data storage library in communication with a host computer and in communication with a second data storage library, wherein said first data storage library comprises a first RAID controller and a first RAID array, wherein said first RAID controller is capable of executing a nominal number of instructions per second (“

NNIPS”

) and is in communication with a second RAID controller disposed in a second data storage library comprising a second RAID array, comprising the steps of;

establishing a normal operation resource allocation;

receiving host I/O requests;

providing information to said second data storage library using a peer-to-peer copy protocol;

operative if there is a data storage device failure in said first RAID array but not in said second RAID array, decreasing a NNIPS allocated for processing host I/O requests.

21. The method of claim 20, further comprising the step of directing said host I/O requests to said second data storage library.

22. The method of claim 20, further comprising the step of decreasing a NNIPS allocated for peer-to-peer copying if there is a data storage device failure in said second RAID array but not in said first RAID array.

23. A RAID controller comprising a processor capable of a nominal number of instructions per second (“

NNIPS”

) and a computer readable medium comprising computer readable program code disposed therein to allocate a percentage of said NNIPS to processing host I/O requests and to peer-to-peer copying, wherein said RAID controller is in communication with a host computer, a first RAID array, and a second RAID controller disposed in a second data storage library comprising a second RAID array, the computer readable program code comprising a series of computer readable program steps to effect;

receiving host I/O requests;

providing information to said second data storage library using a peer-to-peer copy protocol;

operative if there is a data storage device failure in said first RAID array but not in said second RAID array, decreasing a NNIPS allocated for processing host I/O requests.

24. The RAID controller of claim 23, said computer readable program code further comprising a series of computer readable program steps to effect directing said host I/O requests to said second data storage library.

25. The RAID controller of claim 23, said computer readable program code further comprising a series of computer readable program steps to effect decreasing a NNIPS allocated for peer-to-peer copying if there is a data storage device failure in said second RAID array but not in said first RAID array.