Predictive analytics for storage tiering and caching

Patent metadata

Field	Value
Publication number	US-10061702-B2
Application number	US-201514941298-A
Country	US
Kind code	B2
Filing date	Nov 13, 2015
Priority date	Nov 13, 2015
Publication date	Aug 28, 2018
Grant date	Aug 28, 2018

Abstract

Various embodiments for data management across a multiple-tiered storage organization by a processor. Data operations performed across the multiple-tiered storage organization are analyzed over a period of time sufficient to determine usage patterns of the data. Predictive analytics is applied to the usage patterns. Based on the predictive analytics, segments of the data are moved between the multiple-tiered storage organization according to a determined priority account for available system resources, to optimize storage characteristics of the data in the multiple-tiered storage organization.

First claim

The invention claimed is:

1. A method for data management across a multiple-tiered storage organization by a processor, comprising:
analyzing data operations performed across the multiple-tiered storage organization over a period of time sufficient to determine usage patterns of the data by analyzing data read operations from hosts to predict read patterns of the data, the usage patterns inclusive of at least a time of day and the day of a week the data operations are performed;
applying predictive analytics to the usage patterns using idle resources of the multiple-tiered storage organization, the predictive analytics implemented as a background task operation performed by the idle resources such that the data operations are analyzed and the usage patterns of the data are determined in the background task operation performed by the idle resources; and
based on the predictive analytics, moving segments of the data between the multiple-tiered storage organization according to a determined priority accounting for available system resources; wherein the segments of data are moved according to a predetermined priority corresponding to the predicted read patterns from the hosts, including using the time of day and the day of the week the data operations are performed to predict which of the segments of data will be accessed at a future time during the time of day and the day of the week the data operations were previously performed such that the segments of data are moved to facilitate the data operations from the hosts at the future time during the time of day and day of the week the data operations are predicted to be performed, to optimize storage characteristics of the data in the multiple-tiered storage organization.

2. The method of claim 1 , further including configuring the predetermined priority according to a size of the read patterns.

3. The method of claim 1 , further including:
storing at least one of an original storage performance identifier, an original Logical Block Address (LBA) range, and a new location for moved segments of the data, and
examining at least one of the original storage performance identifier, original LBA range, and new location when determining whether the moved segments of the data should be further moved within the multiple-tiered storage organization so as to reduce unnecessary data caching or tiering operations.

4. The method of claim 1 , further including tracking at least one of a read throughput and a write throughput within the multiple-tiered storage organization as a portion of the predictive analytics operations.

5. The method of claim 1 , further including tracking at least one of a read Input/Output Operation per Second (IOPS) and a write IOPS in the multiple-tiered storage organization as a portion of the predictive analytics operations.

6. The method of claim 1 , further including moving the segments of the data according to a user-configurable threshold in at least one of a pre-defined time window and predefined performance window.

7. A system for data management across a multiple-tiered storage organization, comprising:
a processor, operable in the multiple-tiered storage organization, wherein the processor:
analyzes data operations performed across the multiple-tiered storage organization over a period of time sufficient to determine usage patterns of the data by analyzing data read operations from hosts to predict read patterns of the data, the usage patterns inclusive of at least a time of day and the day of a week the data operations are performed;
applies predictive analytics to the usage patterns using idle resources of the multiple-tiered storage organization, the predictive analytics implemented as a background task operation performed by the idle resources such that the data operations are analyzed and the usage patterns of the data are determined in the background task operation performed by the idle resources; and
based on the predictive analytics, moves segments of the data between the multiple-tiered storage organization according to a determined priority accounting for available system resources; wherein the segments of data are moved according to a predetermined priority corresponding to the predicted read patterns from the hosts, including using the time of day and the day of the week the data operations are performed to predict which of the segments of data will be accessed at a future time during the time of day and the day of the week the data operations were previously performed such that the segments of data are moved to facilitate the data operations from the hosts at the future time during the time of day and day of the week the data operations are predicted to be performed, to optimize storage characteristics of the data in the multiple-tiered storage organization.

8. The system of claim 7 , wherein the processor configures the predetermined priority according to a size of the read patterns.

9. The system of claim 7 , wherein the processor:
stores at least one of an original storage performance identifier, an original Logical Block Address (LBA) range, and a new location for moved segments of the data, and
examines at least one of the original storage performance identifier, original LBA range, and new location when determining whether the moved segments of the data should be further moved within the multiple-tiered storage organization so as to reduce unnecessary data caching or tiering operations.

10. The system of claim 7 , wherein the processor tracks at least one of a read throughput and a write throughput within the multiple-tiered storage organization as a portion of the predictive analytics operations.

11. The system of claim 7 , wherein the processor tracks at least one of a read Input/Output Operation per Second (IOPS) and a write IOPS in the multiple-tiered storage organization as a portion of the predictive analytics operations.

12. The system of claim 7 , wherein the processor moves the segments of the data according to a user-configurable threshold in at least one of a pre-defined time window and predefined performance window.

13. A computer program product for data management across a multiple-tiered storage organization by a processor, the computer program product comprising a non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising:
an executable portion that analyzes data operations performed across the multiple-tiered storage organization over a period of time sufficient to determine usage patterns of the data by analyzing data read operations from hosts to predict read patterns of the data, the usage patterns inclusive of at least a time of day and the day of a week the data operations are performed;
an executable portion that applies predictive analytics to the usage patterns using idle resources of the multiple-tiered storage organization, the predictive analytics implemented as a background task operation performed by the idle resources such that the data operations are analyzed and the usage patterns of the data are determined in the background task operation performed by the idle resources; and
an executable portion that, based on the predictive analytics, moves segments of the data between the multiple-tiered storage organization according to a determined priority accounting for available system resources; wherein the segments of data are moved according to a predetermined priority corresponding to the predicted read patterns from the hosts, including using the time of day and the day of the week the data operations are performed to predict which of the s

Assignees

• IBM

Inventors

• Barajas Gonzalez Emmanuel
• Harrington Shaun E
• Mcgregor Harry
• Moore Christopher B

Classifications

• G06F12/0862 Primary

  with prefetch · CPC title

• G06F2212/602

  Details relating to cache prefetching · CPC title

• G06F12/08

  in hierarchically structured memory systems, e.g. virtual memory systems · CPC title

• G06F2212/6024

  History based prefetching · CPC title

• G06F12/0868

  Data transfer between cache memory and other subsystems, e.g. storage devices or host systems · CPC title