Online flash resource migration, allocation, retire and replacement manager based on a cost of ownership model

• US 10,248,348 B2
• Filed: 04/08/2016
• Issued: 04/02/2019
• Est. Priority Date: 08/14/2015
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Yang, Zhengyu, Awasthi, Manu, Ghosh, Mrinmoy, Balakrishnan, Vijay
• Primary Examiner(s): Thai, Tuan V
• Assistant Examiner(s): Waddy, Jr., Edward
• Application Number:
  US15/094,971
• Publication Number:
  US 20170046098A1
• Time in Patent Office: 1,089 Days
• Field of Search: 711154, 711103
• CPC Class Codes: G06F 12/00 Accessing, addressing or allocating within memory systems or architectures digital input from, or digital output to record carriers, e.g. to disk storage units, G06F3/06
• CPC Class Codes: G06F 3/0616 in relation to life time, e.g. increasing Mean Time Between Failures [MTBF]
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/0631 by allocating resources to storage systems
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/0653 Monitoring storage devices or systems
• CPC Class Codes: G06F 3/0688 Non-volatile semiconductor memory arrays
• CPC Class Codes: G06Q 10/06315 Needs-based resource requirements planning or analysis

Abstract

A method for migrating disks includes: dividing a disk pool including a plurality of disks into a random zone and a sequential zone based on sequentiality and randomness of workloads running on the plurality of disks; monitoring a status of each disk in the disk pool based on a total cost of ownership (TCO); migrating one or more workloads of an overheated disk to an idle disk based on the status of each disk. The overheated disk has a first TCO higher than a migration threshold, and the idle disk has a second TCO lower than an idling threshold.

Claims

1. A method comprising:

dividing a disk pool including a plurality of disks into a random zone and a sequential zone based on workload sequentiality and randomness of workloads running on the plurality of disks;

monitoring a status of each disk in the disk pool based on a total cost of ownership (TCO) model that calculates a plurality of cost factors for operating and maintaining the each disk in the disk pool including at least a write amplification factor (WAF) and a replacement cost of the each disk;

identifying an overheated disk and an idle disk among the plurality of disks based on the status of each disk in the disk pool, wherein the overheated disk has a first TCO that is higher than a migration threshold indicating that the overheated disk has a higher degree of write intensive workloads, wherein the idle disk has a second TCO that is lower than an idling threshold, and wherein the idling threshold is lower than the migration threshold; and

migrating one or more workloads of the overheated disk to the idle disk based on the status of each disk and reducing the total cost of ownership for operating the plurality of disks in the disk pool.

2. The method of claim 1, wherein the one or more workloads of the overheated disk are migrated to the idle disk in a lazy mode when a new workload arrives.

3. The method of claim 1, further comprising:

detecting the idle disk in a first zone selected from the random zone and the sequential zone;

migrating one or more workloads of the idle disk to one or more other disks in the first zone;

converting the idle disk from the first zone to a second zone of the random zone and the sequential zone that is different from the first zone; and

migrating the one or more workloads of the overheated disk in the second zone to the idle disk that is converted from the first zone to the second zone.

4. The method of claim 3, wherein the one or more workloads of the overheated disk are migrated to the idle disk in a greedy mode.

5. The method of claim 1, wherein the one or more workloads of the overheated disk are migrated to the idle disk until the TCO of the overheated disk is lowered below the migration threshold.

6. The method of claim 1, wherein the one or more workloads of the overheated disk are migrated to the idle disk until the TCO of the idle disk is higher than the idling threshold.

7. The method of claim 1, further comprising:

detecting a retiring disk that has a TCO higher than a retirement threshold that indicates that a cost of continually operating the retiring disk exceeds a cost of replacing the retiring disk with a replacement disk and operating the replacement disk;

adding the replacement disk to a zone where the retiring disk is detected;

migrating all workloads of the retiring disk to the replacement disk; and

retiring the retiring disk from the zone.

8. The method of claim 7, wherein the retirement threshold is higher than the migration threshold.

9. The method of claim 3, wherein the overheated disk and the idle disk are located in the same zone.

10. The method of claim 7, further comprising detecting the retiring disk in a background process while new workloads are received.

11. A system comprising:

a plurality of disks in a disk pool, wherein the plurality of disks are divided into a first zone and a second zone based on sequentiality and randomness of workloads running on the plurality of disks;

a disk status detector configured to monitor a status of each disk in the disk pool based on a total cost of ownership (TCO) model that calculates a plurality of cost factors for operating and maintaining the each disk in the disk pool including at least a write amplification factor (WAF) and a replacement cost of the each disk; and

a migration manger configured to identify an overheated disk and an idle disk among the plurality of disks based on the status of each disk in the disk pool and migrate one or more workloads of the overheated disk to the idle disk based on the status of each disk to and reduce the total cost of ownership for operating the plurality of disks in the disk pool,

wherein the overheated disk has a first TCO that is higher than a migration threshold indicating that the overheated disk has a higher degree of write intensive workloads, wherein the idle disk has a second TCO that is lower than an idling threshold, and wherein the idling threshold is lower than the migration threshold.

12. The system of claim 11, wherein the one or more workloads of the overheated disk are migrated to the idle disk in a lazy mode when a new workload arrives.

13. The system of claim 11, wherein the disk status detector further is configured to detect the idle disk in the first zone, and

wherein the migration manage is further configured to:

migrate one or more workloads of the idle disk to one or more other disks in the first zone;

convert the idle disk from the first zone to a second zone; and

migrate the one or more workloads of the overheated disk in the second zone to the idle disk that is converted from the first zone to the second zone.

14. The system of claim 13, wherein the one or more workloads of the overheated disk are migrated to the idle disk in a greedy mode.

15. The system of claim 11, wherein the one or more workloads of the overheated disk are migrated to the idle disk until the TCO of the overheated disk is lowered below the migration threshold.

16. The system of claim 11, wherein the one or more workloads of the overheated disk are migrated to the idle disk until the TCO of the idle disk is higher than the idling threshold.

17. The system of claim 11 further comprising a retirement and replacement manager that is configured to:

detect a retiring disk that has a TCO higher than a retirement threshold that indicates that a cost of continually operating the retiring disk exceeds a cost of replacing the retiring disk with a replacement disk and operating the replacement disk;

add the replacement disk to a zone where the retiring disk is detected;

migrate all workloads of the retiring disk to the replacement disk; and

retire the retiring disk from the zone.

18. The system of claim 17, wherein the retirement threshold is higher than the migration threshold.

19. The system of claim 13, the overheated disk and the idle disk are located in the same zone.

20. The system of claim 17, wherein the retirement and replacement manager is further configured to detect the retiring disk in a background process while new workloads are received.