WORKLOAD DETECTION AND MEDIA CACHE MANAGEMENT

• US 20170255556A1
• Filed: 03/02/2016
• Published: 09/07/2017
• Est. Priority Date: 03/02/2016
• Status: Active Grant

• Current Assignee: Seagate Technology LLC (Seagate Technology Holdings Plc)
• Original Assignee: Seagate Technology LLC (Seagate Technology Holdings Plc)
• Inventors: Peng, CheeHou, Bay, PohGuat, Ye, HaiBo, Tang, KayHee
• Granted Patent: US 10,176,119 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 12/126 with special data handling, e.g. priority of data or instructions, handling errors or pinning
• CPC Class Codes: G06F 2212/455 Image or video data
• CPC Class Codes: G06F 2212/464 Multimedia object, e.g. image, video

Abstract

Implementations disclosed herein include a method comprising detecting a workload request from a host, estimating a media cache fill-up rate based on the detected workload request, estimating a current media cache usage, predicting, based on the detected workload request, the estimated media cache fill-up rate and the estimated current media cache usage, a workload profile, and determining a preemptive media cache cleaning strategy based on the predicted workload profile.

Claims

1. A method comprising:

detecting a workload request from a host;

estimating a media cache fill-up rate based on the detected workload request;

estimating a current media cache usage;

predicting, based on the estimated media cache fill-up rate and the estimated current media cache usage, a workload profile; and

determining a preemptive media cache cleaning strategy based on the predicted workload profile.

2. The method of claim 1, wherein the workload request comprises random write commands.

3. The method of claim 1, wherein the workload request comprises random read commands.

4. The method of claim 1, wherein the workload request comprises a combination of random read commands and random write commands.

5. The method of claim 1, wherein the workload request comprises sequential write commands.

6. The method of claim 1, wherein the workload request comprises overlap write commands.

7. The method of claim 1, wherein the preemptive media cache cleaning strategy includes aggressive cleaning.

8. The method of claim 1, wherein the preemptive media cache cleaning strategy includes idle cleaning.

9. The method of claim 1, wherein the preemptive media cache cleaning strategy includes conservative cleaning.

10. The method of claim 1, wherein the preemptive media cache cleaning strategy includes moderate cleaning.

11. A system comprising:

a storage media;

a workload detector configured to detect a workload request from a host;

a resources detector configured to estimate a media cache fill-up rate and a current media cache usage;

a storage controller configured to predict, based on the estimated media cache fill-up rate and the estimated current media cache usage, a workload profile; and

the storage controller further configured to determine a preemptive media cache cleaning strategy based on the predicted workload profile.

12. The system of claim 11, wherein the workload request is one of random write commands, random read commands, a combination of random read commands and random write commands, sequential write commands, and overlap write commands.

13. The system of claim 11, wherein the preemptive media cache cleaning strategy includes one of aggressive cleaning, idle cleaning, conservative cleaning, and moderate cleaning.

14. The system of claim 11, further comprising a table configured to correlate the predicted workload profile with the preemptive media cache cleaning strategy.

15. One or more computer-readable storage media encoding computer-executable instructions for executing on a computer system a computer process that reads only used shingled data tracks adjacent to a target track location, the computer process comprising:

detecting a workload request from a host;

estimating a media cache fill-up rate based on the detected workload request;

estimating a current media cache usage;

predicting, based on the estimated media cache fill-up rate and the estimated current media cache usage, a workload profile; and

determining a preemptive media cache cleaning strategy based on the predicted workload profile.

16. The one or more computer-readable storage media of claim 15, wherein the workload request comprises at least one of random write commands, random write commands, a combination of random read commands and random write commands, sequential write commands, and overlap write commands.

17. The one or more computer-readable storage media of claim 15, wherein the preemptive media cache cleaning strategy further comprising:

determining the workload request to be massive random read and write (mrrw) commands; and

moderately delaying the mrrw command by a predetermined throttle time period prior to processing the mrrw command.

18. The one or more computer-readable storage media of claim 17, wherein the predetermined throttle time period is substantially equal to 32 ms.

19. The one or more computer-readable storage media of claim 16, wherein the the preemptive media cache cleaning strategy further comprising:

determining the workload request to be a massive random write (mrw) command; and

aggressively delaying the mrw command by a predetermined throttle time period prior to processing the mrw command.

20. The one or more computer-readable storage media of claim 23, wherein the predetermined throttle time period is substantially equal to 16 ms.