Granted Patent B2

US 10,997,081 · App. 16/425,455 · Granted May 4, 2021

Inventors: Ramanathan Muthiah (Bangalore, IN); Ramkumar Ramamurthy (Bangalore, IN)

Assignee: Western Digital Technologies, Inc.

G06F12/0873G06F12/10G06F12/12G06F9/345G06F2212/1016G06F2212/1036G06F2212/7201

Abstract

Claims (47)

1. A host comprising:

a controller configured to:

analyze a history of a plurality of prior write commands sent from the host to a storage system to determine whether the plurality of prior write commands comprise a majority of sequential write commands or a majority of random write commands;

determine whether a granularity of a logical-to-physical address map used by the storage system is optimized for sequential write commands or for random write commands;

in response to determining that the history of the plurality of prior write commands does not match the granularity of the logical-to-physical address map used by the storage system, direct the storage system to change the granularity of the logical-to-physical address map used by the storage system to a different granularity; and

send a new write command to the storage system, wherein the different granularity does not necessarily match the new write command because the different granularity was changed based on the history of the prior plurality of write commands and not based on the new write command.

2. The host of claim 1 , wherein changing the granularity of the logical-to-physical address map comprises making the logical-to-physical address map coarser.

3. The host of claim 2 , wherein making the logical-to-physical address map coarser comprises:

consolidating host data;

managing new write entries as a delta file until a coarser logical-to-physical address map is complete; and

flushing the coarser logical-to-physical address map to a memory of the storage system.

4. The host of claim 1 , wherein changing the granularity of the logical-to-physical address map comprises making the logical-to-physical address map finer.

5. The host of claim 4 , wherein making the logical-to-physical address map finer comprises:

splitting the logical-to-physical address map into finer entries;

managing new write entries as a delta file until the splitting is complete; and

flushing the split logical-to-physical address map to a memory of the storage system.

6. The host of claim 1 , wherein the controller is further configured to analyze the history in response to a number of the prior write commands exceeding a threshold and/or in response to a positive health determination of a memory of the storage system.

7. The host of claim 1 , wherein the controller is further configured to analyze the history in runtime across a plurality of write cycles.

8. The host of claim 1 , wherein the controller is further configured to receive an indication from the storage system that the storage system rejected the direction from the host to change the granularity of the logical-to-physical address map used by the storage system.

9. The host of claim 1 , wherein a memory of the storage system comprises a three-dimensional memory.

10. The host of claim 1 , wherein the storage system is configured to be removably connected to the host.

11. A method for storage system calibration, the method comprising:

performing the following in a host in communication with a storage system comprising a memory:

analyzing a plurality of write commands sent to the storage system to determine whether the plurality of prior write commands comprise a majority of sequential write commands or a majority of random write commands;

determining whether a media management layer used by the storage system is optimized for sequential write commands or for random write commands;

in response to determining that the media management layer used by the storage system is optimized for random write commands but that the plurality of prior write commands comprise a majority of sequential write commands, instructing the storage system to make translation mapping performed by the media management layer coarser;

in response to determining that the media management layer used by the storage system results is optimized for sequential write commands but that the plurality of prior write commands comprise a majority of random write commands, instructing the storage system to make translation mapping performed by the media management layer finer; and

sending a new write command to the storage system, wherein the instructed granularity does not necessarily match the new write command because the instructed granularity was changed based on the history of the prior plurality of write commands and not based on the new write command.

12. The method of claim 11 , wherein making translation mapping performed by the media management layer coarser comprises:

consolidating host data;

managing new write entries as a delta file until a coarser media management layer is complete; and

flushing the coarser media management layer to the memory.

13. The method of claim 11 , wherein making translation mapping performed by the media management layer finer comprises:

splitting the media management layer into finer entries;

managing new write entries as a delta file until the splitting is complete; and

flushing the split media management layer to the memory.

14. The method of claim 11 , further comprising determining that the plurality of write commands exceeds a threshold and/or making a positive health determination of the memory.

15. The method of claim 11 , further comprising receiving an indication from the storage system that the storage system rejected the instruction from the host to make the media management layer finer.

16. The method of claim 11 , wherein the analyzing is performed in runtime across a plurality of write cycles.

17. A host comprising:

means for analyzing a history of a plurality of prior write commands sent from the host to a storage system to determine whether the plurality of prior write commands comprise a majority of sequential write commands or a majority of random write commands;

means for determining whether a granularity of a logical-to-physical address map used by the storage system is optimized for sequential write commands or for random write commands;

means for in response to determining that the history of the plurality of prior write commands does not match the granularity of the logical-to-physical address map used by the storage system, directing the storage system to change the granularity of the logical-to-physical address map used by the storage system to a different granularity; and

means for sending a new write command to the storage system, wherein the different granularity does not necessarily match the new write command because the different granularity was changed based on the history of the prior plurality of write commands and not based on the new write command.

18. The host of claim 17 , further comprising means for making the logical-to-physical address map coarser or finer.

19. The host of claim 17 , further comprising means for receiving an indication from the storage system that the storage system rejected the direction from the host to change the granularity of the logical-to-physical address map used by the storage system.

20. The host of claim 17 , wherein the analysis is performed in runtime across a plurality of write cycles.

Assignments (10)