Solid-state disk (SSD) management

• US 9,870,159 B2
• Filed: 10/16/2015
• Issued: 01/16/2018
• Est. Priority Date: 11/02/2010
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Inventors: Cepulis, Darren J.
• Primary Examiner(s): Sadler, Nathan
• Application Number:
  US14/885,617
• Publication Number:
  US 20160034208A1
• Time in Patent Office: 823 Days
• Field of Search: 711103
• CPC Class Codes: G06F 12/0246 in block erasable memory, e.g. flash memory
• CPC Class Codes: G06F 2206/1012 Load balancing
• CPC Class Codes: G06F 2212/7211 Wear leveling
• CPC Class Codes: G06F 3/0604 Improving or facilitating administration, e.g. storage management
• CPC Class Codes: G06F 3/0616 in relation to life time, e.g. increasing Mean Time Between Failures [MTBF]
• CPC Class Codes: G06F 3/0635 by changing the path, e.g. traffic rerouting, path reconfiguration
• CPC Class Codes: G06F 3/0644 Management of space entities, e.g. partitions, extents, pools
• CPC Class Codes: G06F 3/0653 Monitoring storage devices or systems
• CPC Class Codes: G06F 3/0679 Non-volatile semiconductor memory device, e.g. flash memory, one time programmable memory [OTP]
• CPC Class Codes: G06F 3/0688 Non-volatile semiconductor memory arrays
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: G11C 16/3495 Circuits or methods to detect or delay wearout of nonvolatile EPROM or EEPROM memory devices, e.g. by counting numbers of erase or reprogram cycles, by using multiple memory areas serially or cyclically

Abstract

SSD wear-level data (320) is generated on managed nodes (202) having SSDs (206). The wear-level data is collected by a management node (204).

Claims

1. A method comprising:

determining solid-state disk (SSD) wear-level data for SSDs on managed nodes, wherein the wear-level data includes frequency of writes;

determining a recommendation to replace at least one of the SSDs with a hard disk or a replacement SSD based on the frequency of writes, wherein if performance is determined to be less than maximal, the recommendation is to replace the at least one SSD with the hard drive;

otherwise the recommendation is to replace the at least one SSD with the replacement SSD; and

balancing, by a management node, loads among the managed nodes based on the wear-level data.

2. The method of claim 1, comprising:

storing the SSD wear-level data in a database of SSD wear data, wherein the stored SSD wear-level data comprises an identifier for each managed node and the wear-level data for each managed node, and the wear-level data is associated with the identifier of the corresponding managed node.

3. The method of claim 1, wherein balancing loads comprises:

determining workload redistributions among the SSDs so as to stagger times at which each SSD is replaced.

4. The method of claim 1, wherein determining SSD wear-level data comprises:

receiving, at one of the managed nodes, the wear-level data from each of the other managed nodes, wherein the wear-level data of each of the SSDs is measured at the managed node including the SSD.

5. The method of claim 1, wherein determining SSD wear-level data comprises:

receiving, at the management node, the wear-level data for each of the SSDs, wherein the wear-level data of each of the SSDs is measured at the managed node including the SSD.

6. A management node comprising:

a processor to;

receive SSD wear-level data for each SSD of a plurality of SSDs on nodes, wherein the SSD wear-level data for each SSD includes frequency of writes;

compute wear-rate based on the frequency of writes for each SSD from the SSD wear-level data for the SSD, wherein the wear-rate is a rate of change of wear calculated based on previous SSD wear-level data and current wear-level data determined from the SSD wear-level data;

determining a recommendation to replace at least one of the SSDs with a hard disk or a replacement SSD based on the frequency of writes, wherein if performance is determined to be less than maximal, the recommendation is to replace the at least one SSD with the hard drive;

otherwise the recommendation is to replace the at least one SSD with the replacement SSD; and

redistribute loads among the nodes based on wear-rate of each SSD of the plurality of SSDs to extend the lifetimes of some SSDs.

7. The management node of claim 6, wherein the management node redistributes the load to control rates at which the plurality of SSDs are written to.

8. The management node of claim 7, wherein the management node redistributes the load to balance the rates at which the plurality of SSDs are written to.

9. The management node of claim 7, wherein to facilitate the management node redistributes the load to extend remaining lifetime of the plurality of SSDs and controls the rates at which the plurality of SSDs are written to based on the determined redistribution.

10. A computer system comprising:

a processor; and

a non-transitory computer readable medium storing machine readable instructions executable by the processor to;

collect SSD wear-level data for each of a plurality of SSDs from managed nodes having the plurality of SSDs, wherein the SSD wear-level data for each SSD includes frequency of writes;

compute wear-rate for each SSD based on the frequency of writes from the SSD wear-level data for the SSD, wherein the wear-rate is a rate of change of wear calculated based on previous SSD wear-level data and current wear-level data determined from the SSD wear-level data; and

generate a recommendation to replace an SSD of the plurality of SSDs with a hard disk or a replacement SSD based on the wear-rate for the SSD, wherein if performance is less than a maximal value determined based on the wear-rate for the SSD, the recommendation is to replace the SSD with the hard disk;

otherwise the recommendation is to replace the SSD with the replacement SSD.

11. The computer system of claim 10, wherein the processor is to balance loads among the SSDs based on the wear-rate of each of the SSDs.