USING A MACHINE LEARNING MODULE TO PERFORM DESTAGES OF TRACKS WITH HOLES IN A STORAGE SYSTEM

• US 20200133856A1
• Filed: 10/25/2018
• Published: 04/30/2020
• Est. Priority Date: 10/25/2018
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Gupta, Lokesh M., Anderson, Kyler A., Ash, Kevin J., Borlick, Matthew G.
• Granted Patent: US 10,901,904 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 12/0862 with prefetch
• CPC Class Codes: G06F 12/0868 Data transfer between cache memory and other subsystems, e.g. storage devices or host systems
• CPC Class Codes: G06F 12/0871 Allocation or management of cache space
• CPC Class Codes: G06F 2212/1016 Performance improvement
• CPC Class Codes: G06F 2212/154 Networked environment
• CPC Class Codes: G06F 2212/262 configured as RAID
• CPC Class Codes: G06F 2212/312 In storage controller
• CPC Class Codes: G06F 2212/462 Track or segment
• CPC Class Codes: G06F 2212/502 using adaptive policy
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 3/084 Backpropagation, e.g. using gradient descent

Abstract

In response to an end of track access for a track in a cache, a determination is made as to whether the track has modified data and whether the track has one or more holes. In response to determining that the track has modified data and the track has one or more holes, an input on a plurality of attributes of a computing environment in which the track is processed is provided to a machine learning module to produce an output value. A determination is made as to whether the output value indicates whether one or more holes are to be filled in the track. In response to determining that the output value indicates that one or more holes are to be filled in the track, the track is staged to the cache from a storage drive.

Claims

1. A method, comprising:

in response to an end of track access for a track in a cache, determining whether the track has modified data and whether the track has one or more holes; and

in response to determining that the track has modified data and the track has one or more holes, providing input on a plurality of attributes of a computing environment in which the track is processed to a machine learning module to produce an output value;

determining whether the output value indicates whether one or more holes are to be filled in the track; and

in response to determining that the output value indicates that one or more holes are to be filled in the track, staging the track to the cache from a storage drive.

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

in response to completion of staging of the track to the cache from the storage drive, destaging the track from the cache.

3. The method of claim 2, wherein the computing environment comprises a storage controller having the cache, wherein the storage controller is coupled to one or more storage drives in a RAID configuration that stores parity information, wherein the storage controller manages the one or more storage drives to allow input/output (I/O) access to one or more host computing systems.

4. The method of claim 1, wherein the plurality of attributes further includes:

a measure of a current bandwidth of an adapter for the storage drive; and

a measure of an optimum adapter bandwidth of the adapter for the storage drive, wherein the adapter starts thrashing if the optimum adapter bandwidth is exceeded.

5. The method of claim 4, the plurality of attributes further include:

a measure of a speed for a storage rank in which the storage drive is included;

a measure of a response time of the storage rank for staging; and

a measure of a response time of the storage rank for destaging.

6. The method of claim 5, the plurality of attributes include:

a measure of how many task control blocks are allocated for staging;

a measure of how many task control blocks are allocated for destaging;

a measure of how many holes are present in the track; and

a measure of how many requests are queued for staging.

7. The method of claim 6, the plurality of attributes further include:

a measure of an overall non-volatile storage usage;

a measure of a non-volatile storage usage for a rank that includes the storage drive; and

a measure of a maximum non-volatile storage allowed for the rank.

8. A system, comprising:

a memory; and

a processor coupled to the memory, wherein the processor performs operations, the operations comprising;

in response to an end of track access for a track in a cache, determining whether the track has modified data and whether the track has one or more holes; and

in response to determining that the track has modified data and the track has one or more holes, providing input on a plurality of attributes of a computing environment in which the track is processed to a machine learning module to produce an output value;

determining whether the output value indicates whether one or more holes are to be filled in the track; and

in response to determining that the output value indicates that one or more holes are to be filled in the track, staging the track to the cache from a storage drive.

9. The system of claim 8, the operations further comprising:

in response to completion of staging of the track to the cache from the storage drive, destaging the track from the cache.

10. The system of claim 9, wherein the computing environment comprises a storage controller having the cache, wherein the storage controller is coupled to one or more storage drives in a RAID configuration that stores parity information, wherein the storage controller manages the one or more storage drives to allow input/output (I/O) access to one or more host computing systems.

11. The system of claim 8, wherein the plurality of attributes further includes:

a measure of a current bandwidth of an adapter for the storage drive; and

a measure of an optimum adapter bandwidth of the adapter for the storage drive, wherein the adapter starts thrashing if the optimum adapter bandwidth is exceeded.

12. The system of claim 11, the plurality of attributes further include:

a measure of a speed for a storage rank in which the storage drive is included;

a measure of a response time of the storage rank for staging; and

a measure of a response time of the storage rank for destaging.

13. The system of claim 12, the plurality of attributes include:

a measure of how many task control blocks are allocated for staging;

a measure of how many task control blocks are allocated for destaging;

a measure of how many holes are present in the track; and

a measure of how many requests are queued for staging.

14. The system of claim 13, the plurality of attributes further include:

a measure of an overall non-volatile storage usage;

a measure of a non-volatile storage usage for a rank that includes the storage drive; and

a measure of a maximum non-volatile storage allowed for the rank.

15. A computer program product, the computer program product comprising a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code configured to perform operations in a computational device, the operations comprising:

in response to an end of track access for a track in a cache, determining whether the track has modified data and whether the track has one or more holes; and

in response to determining that the track has modified data and the track has one or more holes, providing input on a plurality of attributes of a computing environment in which the track is processed to a machine learning module to produce an output value;

determining whether the output value indicates whether one or more holes are to be filled in the track; and

in response to determining that the output value indicates that one or more holes are to be filled in the track, staging the track to the cache from a storage drive.

16. The computer program product of claim 15, the operations further comprising:

in response to completion of staging of the track to the cache from the storage drive, destaging the track from the cache.

17. The computer program product of claim 16, wherein the computing environment comprises a storage controller having the cache, wherein the storage controller is coupled to one or more storage drives in a RAID configuration that stores parity information, wherein the storage controller manages the one or more storage drives to allow input/output (I/O) access to one or more host computing systems.

18. The computer program product of claim 15, wherein the plurality of attributes further includes:

a measure of a current bandwidth of an adapter for the storage drive; and

a measure of an optimum adapter bandwidth of the adapter for the storage drive, wherein the adapter starts thrashing if the optimum adapter bandwidth is exceeded.

19. The computer program product of claim 18, the plurality of attributes further include:

a measure of a speed for a storage rank in which the storage drive is included;

a measure of a response time of the storage rank for staging; and

a measure of a response time of the storage rank for destaging.

20. The computer program product of claim 19, the plurality of attributes include:

a measure of how many task control blocks are allocated for staging;

a measure of how many task control blocks are allocated for destaging;

a measure of how many holes are present in the track; and

a measure of how many requests are queued for staging.

21. The computer program product of claim 20, the plurality of attributes further include:

a measure of an overall non-volatile storage usage;

a measure of a non-volatile storage usage for a rank that includes the storage drive; and

a measure of a maximum non-volatile storage allowed for the rank.

22. A storage controller, wherein the storage controller is configured to perform operations, the operation, comprising:

in response to an end of track access for a track in a cache, determining whether the track has modified data and whether the track has one or more holes; and

in response to determining that the track has modified data and the track has one or more holes, providing input on a plurality of attributes of a computing environment in which the track is processed to a machine learning module to produce an output value;

determining whether the output value indicates whether one or more holes are to be filled in the track; and

in response to determining that the output value indicates that one or more holes are to be filled in the track, staging the track to the cache from a storage drive.

23. The storage controller of claim 22, the operations further comprising:

in response to completion of staging of the track to the cache from the storage drive, destaging the track from the cache.

24. The storage controller of claim 23, wherein the computing environment comprises a storage controller having the cache, wherein the storage controller is coupled to one or more storage drives in a RAID configuration that stores parity information, wherein the storage controller manages the one or more storage drives to allow input/output (I/O) access to one or more host computing systems.

25. The storage controller of claim 22, wherein the plurality of attributes further includes:

a measure of a current bandwidth of an adapter for the storage drive; and

a measure of an optimum adapter bandwidth of the adapter for the storage drive, wherein the adapter starts thrashing if the optimum adapter bandwidth is exceeded.