Analytics-driven direction for computer storage subsystem device behavior

1. A computer implemented method of directing activity within a computer storage subsystem, comprising:

identifying a computer operating environment including a computer, and a storage subsystem operatively connected to a plurality of storage devices;

receiving by said computer, from a metadata source, metadata representing current and historic performance metrics of said computer operating environment;

identifying by said computer, within the plurality of devices, a first device associated with a first behavior profile governed by a power law distribution, and a second device associated with a second behavior profile governed by a normal distribution;

training, by said computer, a first Machine Learning (ML) model and a second ML model each based, respectively, on said first and second behavior profiles;

establishing, by said computer, a set of Device Performance Rules (DPRs) based on said ML models;

forecasting, by said computer, time-based storage system requirements based, at least in part on said DPRs; and

prefetching, by said computer, data to a cache component based, at least in part on said forecasted system requirements, in accordance with a time reference available to said computer.

2. The method of claim 1 , further including grouping said storage devices into a plurality of application zones based on, at least in part, a common application associated therewith;

establishing, by said computer, a corresponding plurality of zone access protocols based, at least in part on said DPRs; and

guiding activity of said storage devices based, at least in part, on said zone access protocols.

3. The method of claim 1 , wherein said DPRs are based on an ML ensemble model that considers each of said ML models.

4. The method of claim 3 , wherein said ML ensemble is generated via a “bagging with random forest” algorithm.

5. The method of claim 1 , wherein said power law is a Pareto distribution.

6. The method of claim 2 , wherein said guiding of said storage device behavior is carried out, at least in part, by a Field Programmable Gate Array (FPGA) within a core layer of a processor associated with said storage subsystem.

7. The method of claim 1 , wherein said first device is additionally associated with said second behavior protocol.

8. The method of claim 1 , wherein said computer determines said behavior profiles by analyzing said metadata.

9. A system to direct activity within a computer storage subsystem, which comprises:

a computer system comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to:

identify a computer operating environment including a computer, and a storage subsystem operatively connected to a plurality of storage devices;

receive, from a metadata source, metadata representing current and historic performance metrics of said computer operating environment;

identify, within the plurality of devices, a first device associated with a first behavior profile governed by a power law distribution, and a second device associated with a second behavior profile governed by a normal distribution;

train a first Machine Learning (ML) model and a second ML model each based, respectively, on said first and second behavior profiles;

establish a set of Device Performance Rules (DPRs) based on said ML models;

forecasting, by said computer, time-based storage system requirements based, at least in part on said DPRs; and

prefetching, by said computer, data to a cache component based, at least in part on said forecasted system requirements, in accordance with a time reference available to said computer.

10. The system of claim 9 further including instructions causing said computer to group said storage devices into a plurality of application zones based, at least in part, on a common application associated therewith; establish, a corresponding plurality of zone access protocols based, at least in part, on said DPRs; and guide activity of said storage devices based, at least in part, on said zone access protocols.

11. The system of claim 9 , wherein said DPRs are based on an ML ensemble model that considers each of said ML models.

12. The system of claim 9 , wherein said power law is a Pareto distribution.

13. The system of claim 10 , wherein said guiding of said storage device behavior is carried out, at least in part, by a Field Programmable Gate Array (FPGA) within a core layer of a processor associated with said storage subsystem.

14. The system of claim 9 , wherein said first device is additionally associated with said second behavior protocol.

15. The system of claim 9 , wherein said computer determines said behavior profiles by analyzing said metadata.

16. A computer program product to direct activity within a computer storage subsystem, the program instructions executable by a computer to cause the computer to:

identify, using a computer operating environment including a computer, and a storage subsystem operatively connected to a plurality of storage devices;

receive, using said computer, from a metadata source, metadata representing current and historic performance metrics of said computer operating environment;

identify, using said computer, within the plurality of devices, a first device associated with a first behavior profile governed by a power law distribution, and a second device associated with a second behavior profile governed by a normal distribution;

train, using said computer, a first Machine Learning (ML) model and a second ML model each based, respectively, on said first and second behavior profiles;

establish, using said computer, a set of Device Performance Rules (DPRs) based on said ML models;

forecasting, using said computer, by said computer, time-based storage system requirements based, at least in part on said DPRs; and

prefetching, using said computer, by said computer, data to a cache component based, at least in part on said forecasted system requirements, in accordance with a time reference available to said computer.

17. The computer program product of claim 16 further including instructions causing said computer to group said storage devices into a plurality of application zones based, at least in part, on a common application associated therewith; establish, a corresponding plurality of zone access protocols based, at least in part, on said DPRs; and guide activity of said storage devices based, at least in part, on said zone access protocols.

18. The computer program product of claim 16 , wherein said DPRs are based on an ML ensemble model that considers each of said ML models.

19. The computer program product of claim 17 , wherein said guiding of said storage device behavior is carried out, at least in part, by a Field Programmable Gate Array (FPGA) within a core layer of a processor associated with said storage subsystem.

20. The computer program product of claim 16 , wherein said computer determines said behavior profiles by analyzing said metadata.