Reducing database system query transaction delay

Patent metadata

Field	Value
Publication number	US-2021019321-A1
Application number	US-201916514401-A
Country	US
Kind code	A1
Filing date	Jul 17, 2019
Priority date	Jul 17, 2019
Publication date	Jan 21, 2021
Grant date	—

Abstract

A processing system including at least one processor may obtain a first set of performance records of a database system, train a machine learning model in accordance with the first set of performance records, where the machine learning model that is trained in accordance with the first set of performance records is configured to predict a latency of a query transaction for a designated time period, present a user interface with a plurality of settings of the database system that are user-adjustable, where the plurality of settings is associated with at least a portion of the first set of performance records, calculate a first predicted latency of a query transaction at the designated time period via the machine learning model in accordance with a set of values of the plurality of settings, and present the first predicted latency via the user interface.

First claim

What is claimed is:

1 . A method comprising:
obtaining, by a processing system including at least one processor, a first set of performance records of a database system; training, by the processing system, a machine learning model in accordance with the first set of performance records, where the machine learning model that is trained in accordance with the first set of performance records is configured to predict a query sub-operation delay at a server node of the database system for a designated time period; obtaining, by the processing system via a user interface, at least one input selecting the designated time period; selecting, by the processing system, a set of values of a plurality of configuration settings for at least one resource quota pool of the database system for the designated time period at the server node in accordance with the machine learning model; and presenting, by the processing system, the set of values of the plurality of configuration settings via the user interface.

2 . The method of claim 1 , wherein the training the machine learning model in accordance with the first set of performance records utilizes time stamps and a plurality of performance metrics associated with the first set of performance records as inputs, wherein for each performance record of the first set of performance records, the plurality of performance metrics includes: configuration setting values for a plurality resource quota pools, observed values associated with usage of the database system, and a delay measurement, wherein the plurality of resource quota pools includes the at least one resource quota pool.

3 . The method of claim 1 , wherein the machine learning model comprises:
a plurality of independent variables representing a plurality of performance metrics associated with the first set of performance records of the database system; and at least one dependent variable comprising the query sub-operation delay at the server node of the database system for the designated time period.

4 . The method of claim 3 , wherein the selecting comprises:
identifying a selected number of candidate performance metrics of the plurality of performance metrics with a greatest effect on the query sub-operation delay according to the machine learning model; identifying performance metrics of the candidate performance metrics that are associated with adjustable configuration settings of the database system, wherein the plurality of configuration settings for which the set of values is selected comprises the adjustable configuration settings that are identified; and identifying the set of values of the plurality of configuration settings that minimizes the query sub-operation delay according to the machine learning model.

5 . The method of claim 4 , wherein for the identifying the set of values of the plurality of configuration settings of the adjustable configuration settings that minimizes the query sub-operation delay, performance metrics of the candidate performance metrics that are not associated with the adjustable configuration settings are assumed to be average values based upon the first set of performance records.

6 . The method of claim 1 , further comprising:
obtaining an input to implement the set of values of the plurality of configuration settings for the designated time period at the server node; and sending an instruction to the server node to implement the set of values of the plurality of configuration settings for the designated time period.

7 . The method of claim 1 , further comprising:
obtaining a second set of performance records of the database system; detecting that a deviation of the second set of performance records from the first set of performance records exceeds a threshold deviation; and retraining the machine learning model in accordance with the second set of performance records.

8 . The method of claim 1 , wherein the at least one input is further selecting a given resource quota pool of the at least one resource quota pool, wherein the machine learning model that is trained in accordance with the first set of performance records is configured to predict the query sub-operation delay at the server node of the database system for the given resource quota pool for the designated time period.

9 . The method of claim 8 , wherein the training the machine learning model in accordance with the first set of performance records utilizes time stamps and a plurality of performance metrics associated with the first set of performance records as inputs, wherein for each performance record of the first set of performance records, the plurality of performance metrics includes: a resource quota pool identifier, configuration setting values for a plurality resource quota pools, observed values associated with usage of the database system, and a delay measurement, wherein the plurality of resource quota pools includes the at least one resource quota pool.

10 . The method of claim 8 , wherein the selecting comprises selecting the set of values of the plurality of configuration settings for at least the given resource quota pool for the designated time period at the server node in accordance with the machine learning model.

11 . The method of claim 10 , further comprising:
obtaining an input to implement the set of values of the plurality of configuration settings for at least the given resource quota pool for the designated time period at the server node; and sending an instruction to the server node to implement the set of values of the plurality of configuration settings for at least the given resource quota pool for the designated time period.

12 . The method of claim 8 , wherein the given resource quota pool is associated with a designated type of query sub-operation.

13 . The method of claim 1 , wherein the at least one input is further selecting a particular query sub-operation type, wherein the machine learning model that is trained in accordance with the first set of performance records is configured to predict the query sub-operation delay at the server node of the database system for the particular query sub-operation type for the designated time period.

14 . The method of claim 13 , wherein the training the machine learning model in accordance with the first set of performance records utilizes time stamps and a plurality of performance metrics associated with the first set of performance records as inputs, wherein for each performance record of the first set of performance records, the plurality of performance metrics includes: configuration setting values for a resource quota pool in which a query sub-operation executes, observed values associated with usage of the database system, and a delay measurement.

15 . The method of claim 13 , wherein the selecting comprises selecting the set of values of the plurality of configuration settings for the particular query sub-operation type for the designated time period at the server node in accordance with the machine learning model.

16 . The method of claim 15 , further comprising:
obtaining an input to implement the set of values of the plurality of configuration settings for the particular query sub-operation type for the designated time period at the server node; and sending an instruction to the server node to implement the set of values of the plurality of configuration settings for the particular query sub-operation type for the designated time period.

17 . The method of claim 16 , further comprising:
establishing a new resource quota pool for the particular query su

Assignees

• At & T Ip I Lp

Inventors

• Ehrlich Willa Kay
• Frabetti Jo A
• Lu Yangyi

Classifications

• G06N20/00

  Machine learning · CPC title

• G06F16/24553 Primary

  of query operations · CPC title

• G06F9/5061 Primary

  Partitioning or combining of resources · CPC title

• G06F2209/501

  Performance criteria · CPC title

• G06F2209/5011

  Pool · CPC title