COLLABORATIVE HEALTH MANAGEMENT IN A STORAGE SYSTEM

• US 20150178191A1
• Filed: 12/24/2013
• Published: 06/25/2015
• Est. Priority Date: 12/24/2013
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: CAMP, CHARLES J., WALLS, ANDREW D.
• Granted Patent: US 9,619,381 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 12/0246 in block erasable memory, e.g. flash memory
• CPC Class Codes: G06F 2212/7201 Logical to physical mapping or translation of blocks or pages
• CPC Class Codes: G06F 2212/7211 Wear leveling

Abstract

In at least one embodiment, multiple controllers implement collaborative management of a non-volatile hierarchical storage system. In the storage system, a first controller receives health reports from at least second and third controllers regarding health of multiple storage units of physical storage under control of the second and third controllers and maintains a health database of information received in the health reports. In response to a health event and based on information in the health database, the first controller modifies logical-to-physical address mappings of one or more of multiple storage units under its control such that data having greater access heat is mapped to relatively healthier storage units and data having less access heat is mapped to relatively less healthy storage units. Thereafter, the first controller directs write requests to storage units under its control in accordance with the modified logical-to-physical address mappings.

Claims

1. A method of collaborative management in a non-volatile hierarchical storage system including multiple controllers, the method comprising:

a first controller receiving health reports from at least second and third controllers of the storage system regarding health of multiple storage units of physical storage under control of the second and third controllers;

the first controller maintaining a health database of information received in the health reports;

in response to a health event and based on information in the health database, the first controller modifying logical-to-physical address mappings of one or more of multiple storage units under its control such that data having greater access heat is mapped to relatively healthier storage units and data having less access heat is mapped to relatively less healthy storage units; and

thereafter, the first controller directing write requests to storage units under its control in accordance with the modified logical-to-physical address mappings.

2. The method of claim 1, and further comprising:

in response to the health event, the first controller relocating data stored in at least one of the multiple storage units under its control to another storage unit in the storage system.

3. The method of claim 1, and further comprising:

in response to the health event, the first controller causing an alert to be presented recommending addition to the storage system of a storage unit.

4. The method of claim 1, and further comprising determining an associated health grade for each of the multiple storage units based on at least one error metric for each of the multiple storage units.

5. The method of claim 1, and further comprising:

the first controller determining a combined health grade for multiple storage units under its control based on information received in the health reports; and

the first controller providing a health report regarding multiple storage units under its control to at least one other controller in the storage system.

6. The method of claim 1, wherein the first and second controllers are peer controllers.

7. A first controller for a non-volatile hierarchical storage system, wherein:

the first controller is configured such that the first controller receives health reports from at least second and third controllers of the storage system regarding health of multiple storage units of physical storage under control of the second and third controllers and maintains a health database of information received in the health reports, wherein the first controller, in response to a health event and based on information in the health database, modifies logical-to-physical address mappings of one or more of multiple storage units under its control such that data having greater access heat is mapped to relatively healthier storage units and data having less access heat is mapped to relatively less healthy storage units and thereafter directs write requests to storage units under its control in accordance with the modified logical-to-physical address mappings.

8. The first controller of claim 7, wherein the first controller is further configured such that, in response to the health event, the first controller relocates data stored in at least one of the multiple storage units under its control to another storage unit in the storage system.

9. The first controller of claim 7, wherein the first controller is further configured such that the first controller, responsive to the health event, causes an alert to be presented recommending addition to the storage system of a storage unit.

10. The first controller of claim 7, wherein the first controller is further configured such that the first controller determines an associated health grade for each of the multiple storage units based on at least one error metric for each of the multiple storage units.

11. The first controller of claim 7, wherein the first controller is further configured such that the first controller determines a combined health grade for multiple storage units under its control based on information received in the health reports and provides a health report regarding multiple storage units under its control to at least one other controller in the storage system.

12. The first controller of claim 7, wherein the first controller is a peer controller of the second controller.

13. A storage system, comprising:

the first controller of claim 7;

the second and third controllers communicatively coupled to the first controller; and

the multiple storage units controlled by the first, second and third controllers, wherein at least some of the storage units are hierarchically arranged.

14. The first controller of claim 7, wherein the first controller comprises a redundant array of inexpensive disks (RAID) controller.

15. A program product, comprising:

a storage device; and

program code stored in the storage device, wherein the program code, when executed by a first controller that controls multiple storage devices in a non-volatile hierarchical storage system, causes the first controller to perform;

the first controller receiving health reports from at least second and third controllers of the storage system regarding health of multiple storage units of physical storage under control of the second and third controllers;

the first controller maintaining a health database of information received in the health reports;

in response to a health event and based on information in the health database, the first controller modifying logical-to-physical address mappings of one or more of multiple storage units under its control such that data having greater access heat is mapped to relatively healthier storage units and data having less access heat is mapped to relatively less healthy storage units; and

thereafter, the first controller directing write requests to storage units under its control in accordance with the modified logical-to-physical address mappings.

16. The program product of claim 15, wherein the program code further causes the first controller to perform:

in response to the health event, the first controller relocating data stored in at least one of the multiple storage units under its control to another storage unit in the storage system.

17. The program product of claim 15, wherein the program code further causes the first controller to perform:

in response to the health event, the first controller causing an alert to be presented recommending addition to the storage system of a storage unit.

18. The program product of claim 15, wherein the program code further causes the first controller to perform:

determining an associated health grade for each of the multiple storage units based on at least one error metric for each of the multiple storage units.

19. The program product of claim 15, wherein the program code further causes the first controller to perform:

the first controller determining a combined health grade for multiple storage units under its control based on information received in the health reports; and

the first controller providing a health report regarding multiple storage units under its control to at least one other controller in the storage system.

20. The program product of claim 15, wherein the first and second controllers are peer controllers.