Utilizing machine learning to determine data storage pruning parameters

Patent metadata

Field	Value
Publication number	US-11544217-B2
Application number	US-201916431092-A
Country	US
Kind code	B2
Filing date	Jun 4, 2019
Priority date	Apr 11, 2018
Publication date	Jan 3, 2023
Grant date	Jan 3, 2023

Abstract

A device receives, from a user device, a request to prune a primary database, and receives primary database information associated with the primary database and secondary database information associated with a secondary database that is different than the primary database. The device processes the primary database information and the secondary database information, with a machine learning model, to generate suggested pruning parameters, and provides the suggested pruning parameters to the user device. The device receives selected pruning parameters from the user device, where the selected pruning parameters are selected from the suggested pruning parameters or are input via the user device. The device removes pruned information from the primary database based on the selected pruning parameters, and provides the pruned information to the secondary database based on the selected pruning parameters.

First claim

What is claimed is:

1. A method, comprising:
receiving, by a device, first information identifying a primary database, and second information identifying a secondary database,
wherein the first information includes metadata associated with the primary database;

processing, by the device and by using a machine learning model, the first information and the second information, to generate recommended pruning parameters,
the recommended pruning parameters including at least one of:
a pruning frequency parameter,
a primary database name parameter,
a column to query parameter, or
a date parameter;

receiving, by the device, information associated with selecting one or more pruning parameters selected from the recommended pruning parameters;
removing, by the device, first data from the primary database based on the selected one or more pruning parameters,
wherein the first data is stored on the primary database prior to being removed;

providing, by the device, the first data to the secondary database based on the selected one or more pruning parameters;
determining, by the device and based on a pattern identified from historical access to the first data, a predicted point in time associated with the first data,
the historical access to the first data including at least one of:
data associated with the historical access to the first data when the first data was stored in the primary database, or
data associated with the historical access to the first data when the first data was stored in the secondary database;

determining, by the device, that the first data was removed based on the selected one or more pruning parameters;
removing, by the device and based on determining that the first data was removed based on the selected one or more pruning parameters, the first data from the secondary database prior to the predicted point in time;
providing, by the device and based on determining that the first data was removed based on the selected one or more pruning parameters, the first data to the primary database prior to the predicted point in time;
receiving, by the device and based on an alert associated with at least one of the primary database or the secondary database, reconfigured one or more pruning parameters;
identify, by the device, based on analyzing the metadata associated with the primary database, and based on the reconfigured one or more pruning parameters, second data, that is different from the first data, to be removed from the primary database; and
determining, based on identifying the second data, a periodic schedule for the second data to be removed from the primary database.

2. The method of claim 1 , further comprising:
analyzing the first information to determine a utilization frequency parameter associated with the first data; and
identifying the first data based on the utilization frequency parameter.

3. The method of claim 1 , further comprising:
analyzing other metadata associated with the primary database,
the other metadata comprising data specifying a frequency of use associated with a portion of the first data; and

generating a pruning parameter, of the one or more pruning parameters, based on the frequency of use and analyzing the other metadata.

4. The method of claim 1 , further comprising:
analyzing other metadata associated with the primary database,
the other metadata comprising data specifying a quantity of data associated with a portion of the first data; and

generating the one or more pruning parameters based on the quantity of data and the analysis of the other metadata.

5. The method of claim 1 , wherein the first data includes at least one database record that includes data indicating a previous access time associated with the at least one database record, and
wherein providing the first data to the secondary database based on the one or more pruning parameters comprises:
providing the first data to the secondary database based on the one or more pruning parameters and the data indicating the previous access time.

6. The method of claim 1 , wherein the first data is associated with a first cost associated with storing the first data in the primary database,
wherein the first data is associated with a second cost associated with storing the first data in the secondary database, and
wherein providing the first data to the secondary database based on the one or more pruning parameters comprises:
providing the first data to the secondary database based on the one or more pruning parameters, the first cost, and the second cost.

7. The method of claim 1 , wherein the one or more pruning parameters include at least one of:
a parameter indicating a frequency to prune the primary database,
a parameter indicating information to be pruned,
a parameter indicating a time or date range associated with the information to be pruned, or
a parameter associated with the secondary database.

8. A device, comprising:
one or more memories; and
one or more processors communicatively coupled to the one or more memories, configured to:
receive first information identifying a primary database, and second information identifying a secondary database,
wherein the first information includes metadata associated with the primary database;

process, by using a machine learning model, the first information and the second information, to generate recommended pruning parameters,
the recommended pruning parameters including at least one of:
a pruning frequency parameter,
a primary database name parameter,
a column to query parameter, or
a date parameter;

receive information associated with selecting one or more pruning parameters selected from the recommended pruning parameters;
remove first data from the primary database based on the selected one or more pruning parameters,
wherein the first data is stored on the primary database prior to being removed;

provide the first data to the secondary database based on the selected one or more pruning parameters;
determine, based on a pattern identified from historical access to the first data, a predicted point in time associated with the first data,
the historical access to the first data including at least one of:
data associated with the historical access to the first data when the first data was stored in the primary database, or
data associated with the historical access to the first data when the first data was stored in the secondary database;

determine that the first data was removed based on the selected one or more pruning parameters;
remove, based on determining that the first data was removed based on the selected one or more pruning parameters, the first data from the secondary database prior to the predicted point in time;
provide the first data to the primary database prior to the predicted point in time;
receive, based on an alert associated with at least one of the primary database or the secondary database, reconfigured one or more pruning parameters;
identify, based on analyzing the metadata associated with the primary database, and based on the reconfigured one or more pruning parameters, second data to be removed from the primary database; and
determine, based on identifying the second data, a periodic schedule for the second data to be removed from the primary database.

9. The device of claim 8 , wherein the one or more processors are further configured to:
analyze the first information to determine a utilization frequency parameter associated with the first data; and
identify the first data based on the utilization frequency parameter.

10. The device of claim 8 , wherein the one or more processors are further configured to:
analyze other

Assignees

• Capital One Services Llc

Inventors

• Deshpande Amit
• Subramanian Ponnazhakan
• Chikkaveerappa Satish
• Pandya Asutosh
• Venuraju Mithra Kosur
• Mohammed Sahul

Classifications

• G06F16/2365

  Ensuring data consistency and integrity · CPC title

• G06F16/113 Primary

  Details of archiving (lifecycle management in storage systems G06F3/0649; point-in-time backing up or restoration of persistent data G06F11/1446) · CPC title

• G06F16/284

  Relational databases · CPC title

• G06F16/125 Primary

  characterised by the use of retention policies (retention policies for HSM systems G06F16/185) · CPC title