Determination of redundant array of independent disk level for storage of datasets

Patent metadata

Field	Value
Publication number	US-10528258-B2
Application number	US-201815908114-A
Country	US
Kind code	B2
Filing date	Feb 28, 2018
Priority date	Feb 28, 2018
Publication date	Jan 7, 2020
Grant date	Jan 7, 2020

Abstract

A storage controller receives a write command from a host computational device to write a dataset. A learning application maps the dataset into a Redundant Array of Independent Disk (RAID) level by classifying the dataset into a category of a plurality of categories, wherein the plurality of categories correspond to a plurality of RAID levels. A storage management application writes the dataset to a RAID array that corresponds to the RAID level determined by the learning application.

First claim

What is claimed is:

1. A method, comprising:
receiving, by a storage controller, a write command from a host computational device to write a dataset;
mapping, by a learning application, the dataset into a Redundant Array of Independent Disk (RAID) level by classifying the dataset into a category of a plurality of categories, wherein the plurality of categories correspond to a plurality of RAID levels; and
writing by a storage management application, the dataset to a RAID array that corresponds to the RAID level determined by the learning application, wherein a tiering application that executes in the storage controller moves the dataset from one storage tier to another storage tier overriding a previous classification of the dataset made by the learning application, and wherein the learning application maps the dataset into the RAID level based on indications corresponding to:
an initiator indication that indicates the host computational device from which a request to write the dataset was received by the storage controller;
a block size indication that indicates a size in blocks of the dataset; and
a timestamp indication that indicates a time at which each dataset was generated.

2. The method of claim 1 , wherein the learning application performs operations comprising:
performing a training phase to adjust parameters in the learning application by processing a plurality of training datasets provided for the training phase, wherein the plurality of training datasets provided for the training phase indicate mappings of the plurality of training datasets into a plurality of RAID levels; and
improving classification of datasets while processing newly arriving datasets to map the newly arriving datasets to RAID levels.

3. The method of claim 2 , wherein the learning application interfaces with the tiering application to adjust the classification of datasets, in response to the tiering application moving one or more datasets from one storage tier to another storage tier.

4. The method of claim 3 , wherein the mapping of the dataset into the RAID level is based on fault tolerance requirements and performance requirements of the dataset.

5. The method of claim 1 , wherein a first dataset that requires a higher level of fault tolerance than a second dataset is mapped into a RAID level that provides a higher level of fault tolerance in comparison to a RAID level to which the second dataset is mapped into.

6. The method of claim 1 , wherein the classifying of the dataset to a selected RAID level is based on determining that a probability of the dataset being of the selected RAID level is equal to a multiplication of probabilities of the selected RAID level multiplied by the probability of an incoming initiator indication given that the selected RAID has happened, multiplied by the probability of the block size indication being obtained given that the selected RAID has happened, multiplied by the probability of the timestamp indication given that the selected RAID has happened.

7. The method of claim 1 , wherein if the dataset is written to an incorrect RAID level by the learning application then the tiering application moves the dataset to a correct RAID level to trigger a new learning cycle for classifying the dataset.

8. A system, comprising:
a memory; and
a processor coupled to the memory, wherein the processor performs operations, the operations performed by the processor comprising:
receiving a write command from a host computational device to write a dataset;
mapping, by a learning application, the dataset into a Redundant Array of Independent Disk (RAID) level by classifying the dataset into a category of a plurality of categories, wherein the plurality of categories correspond to a plurality of RAID levels; and
writing by a storage management application, the dataset to a RAID array that corresponds to the RAID level determined by the learning application, wherein a tiering application that executes in the system moves the dataset from one storage tier to another storage tier overriding a previous classification of the dataset made by the learning application, and wherein the learning application maps the dataset into the RAID level based on indications corresponding to:
an initiator indication that indicates the host computational device from which a request to write the dataset was received by the system;
a block size indication that indicates a size in blocks of the dataset; and
a timestamp indication that indicates a time at which each dataset was generated.

9. The system of claim 8 , wherein the learning application performs operations comprising:
performing a training phase to adjust parameters in the learning application by processing a plurality of training datasets provided for the training phase, wherein the plurality of training datasets provided for the training phase indicate mappings of the plurality of training datasets into a plurality of RAID levels; and
improving classification of datasets while processing newly arriving datasets to map the newly arriving datasets to RAID levels.

10. The system of claim 9 , wherein the learning application interfaces with the tiering application to adjust the classification of datasets, in response to the tiering application moving one or more datasets from one storage tier to another storage tier.

11. The system of claim 10 , wherein the mapping of the dataset into the RAID level is based on fault tolerance requirements and performance requirements of the dataset.

12. The system of claim 8 , wherein a first dataset that requires a higher level of fault tolerance than a second dataset is mapped into a RAID level that provides a higher level of fault tolerance in comparison to a RAID level to which the second dataset is mapped into.

13. The system of claim 8 , wherein the classifying of the dataset to a selected RAID level is based on determining that a probability of the dataset being of the selected RAID level is equal to a multiplication of probabilities of the selected RAID level multiplied by the probability of an incoming initiator indication given that the selected RAID has happened, multiplied by the probability of the block size indication being obtained given that the selected RAID has happened, multiplied by the probability of the timestamp indication given that the selected RAID has happened.

14. The system of claim 8 , wherein if the dataset is written to an incorrect RAID level by the learning application then the tiering application moves the dataset to a correct RAID level to trigger a new learning cycle for classifying the dataset.

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 on a storage controller, the operations comprising:
receiving, by the storage controller, a write command from a host computational device to write a dataset;
mapping, by a learning application, the dataset into a Redundant Array of Independent Disk (RAID) level by classifying the dataset into a category of a plurality of categories, wherein the plurality of categories correspond to a plurality of RAID levels; and
writing by a storage management application, the dataset to a RAID array that corresponds to the RAID level determined by the learning application, wherein a tiering application that executes in the storage controller moves the dataset from one storage tier to another storage tier overriding a previous classification of the dataset made by the learning application, and wh

Assignees

• IBM

Inventors

• Barajas Gonzalez Emmanuel
• Harrington Shaun E
• Anna Gary

Classifications

• G06F18/24155

  Bayesian classification · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06F3/0604 Primary

  Improving or facilitating administration, e.g. storage management · CPC title

• G06F3/0614

  Improving the reliability of storage systems · CPC title

• G06F3/0689

  Disk arrays, e.g. RAID, JBOD · CPC title