Message broker resource deletion

Patent metadata

Field	Value
Publication number	US-11917031-B2
Application number	US-202217581369-A
Country	US
Kind code	B2
Filing date	Jan 21, 2022
Priority date	Jan 21, 2022
Publication date	Feb 27, 2024
Grant date	Feb 27, 2024

Abstract

A message broker resource monitoring service obtains message broker resource parameter data of a resource, based on communications of a message broker. Based on the resource parameter data and historical data, the message broker resource monitoring service obtains a predicted message delivery time value, which can be in association with confidence. If the predicted message delivery time value satisfies a resource deletion criterion, e.g., the predicted message delivery time value, with sufficient confidence, exceeds a threshold value, the message broker resource monitoring service triggers an action to delete the resource. To obtain the predicted value, a regression such as symmetric conformal quantile regression can be applied to the parameter data, e.g., to obtain a predicted message delivery time/latency value.

First claim

What is claimed is:

1. A system, comprising:
a processor; and
a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations, the operations comprising:
monitoring communications of a message broker to collect resource parameter data of a resource of the message broker, wherein a behavior of the message broker is representative of a collective behavior of one or more microservices associated with the message broker;
obtaining, based on the resource parameter data of the resource of the message broker and historical data, a predicted message delivery time value;
determining whether the predicted message delivery time value satisfies a resource deletion criterion;
in response to determining that the predicted message delivery time value satisfies the resource deletion criterion, taking an action to delete the resource, wherein deletion of the resource is semi-automatic, resulting in the resource being deleted based on the resource not being utilized within a defined time interval; and
training a regression model used to obtain the predicted message delivery time value based on a number of incoming data points being determined to be equal to a number of the one or more microservices associated with the message broker, wherein a regression technique applied to obtain the predicted message delivery time value comprises symmetric conformal quantile regression.

2. The system of claim 1 , wherein the resource comprises a queue of the message broker.

3. The system of claim 1 , wherein the resource comprises at least one of: a socket, memory space, a file handler, disk space, an open connection, or a channel used by the message broker.

4. The system of claim 1 , wherein the message broker is coupled to a microservice to handle communications of the microservice.

5. The system of claim 1 , wherein the operations further comprise sending heartbeats to the message broker for connection confirmations.

6. The system of claim 1 , wherein the resource parameter data comprises at least one of: measure node data, message rate data, publisher confirmation data, acknowledgement time data, or order of publisher confirmation data.

7. The system of claim 1 , wherein the resource parameter data comprises at least one of: connection data or queue data.

8. The system of claim 1 , wherein the resource deletion criterion comprises a threshold value.

9. The system of claim 1 , wherein the resource deletion criterion comprises a threshold value and a confidence score.

10. The system of claim 9 , wherein the confidence score is between an upper and a lower confidence bound.

11. The system of claim 1 , wherein the taking the action to delete the resource comprises outputting a notification.

12. The system of claim 1 , wherein obtaining the resource parameter data of the resource of the message broker is based on user-configurable time data.

13. A method, comprising:
collecting, by a message broker resource monitoring service executing via a system comprising a processor, a current resource parameter dataset based on communications of a monitored message broker, wherein a behavior of the monitored message broker is representative of a collective behavior of one or more microservices associated with the monitored message broker;
obtaining, by the message broker resource monitoring service, a predicted message delivery latency value based on the current resource parameter dataset and a historical resource parameter dataset based on previous message broker communications;
determining, by the message broker resource monitoring service, whether the predicted message delivery latency value satisfies a resource deletion criterion of a message broker resource;
in response to the predicted message delivery latency value satisfying the resource deletion criterion, triggering, by the message broker resource monitoring service, a deletion of the message broker resource, wherein the deletion is semi-automatic, resulting in the message broker resource being deleted based on the message broker resource not being utilized before expiration of a defined time limit; and
in response to a number of incoming data points being determined to be equal to a number of the one or more microservices associated with the monitored message broker, training a regression model used to obtain the predicted message delivery latency value, wherein a regression technique applied to obtain the predicted message delivery time value comprises symmetric conformal quantile regression.

14. The method of claim 13 , wherein the resource deletion criterion comprises a criterion based on a threshold message delivery latency value, and wherein the regression technique estimates the predicted message delivery latency value.

15. The method of claim 13 , wherein the resource deletion criterion comprises a threshold message delivery latency value and first confidence data, and wherein the predicted message delivery latency value is associated with estimated confidence data.

16. A non-transitory machine-readable medium, comprising executable instructions that, when executed by a processor, facilitate performance of operations, the operations comprising:
collecting resource parameter data of a resource used by message broker of a microservices environment, wherein a behavior of the message broker is representative of a collective behavior of one or more microservices in the microservices environment;
obtaining, based on the resource parameter data and historical data, a predicted message delivery time value and associated prediction confidence data;
determining, based on the predicted message delivery time value and the associated prediction confidence data, that the predicted message delivery time value satisfies a resource deletion criterion;
triggering a deletion of a message broker resource, wherein the deletion is automatic or semi-automatic, and wherein a semi-automatic deletion results in the message broker resource being deleted based on the message broker resource not being utilized within a defined time interval; and
in response to a number of incoming data points being determined to be equal to a number of the one or more microservices associated with the message broker, training a regression model usable to obtain the predicted message delivery time value, wherein a regression technique applied to obtain the predicted message delivery time value comprises symmetric conformal quantile regression.

17. The non-transitory machine-readable medium of claim 16 , wherein the obtaining the predicted message delivery time value and the associated prediction confidence data comprises applying the regression technique based on the resource parameter data and the historical data.

18. The non-transitory machine-readable medium of claim 16 , wherein the resource parameter data comprises at least one of: measure node data, message rate data, publisher confirmation data, acknowledgement time data, or order of publisher confirmation data.

19. The non-transitory machine-readable medium of claim 16 , wherein the resource parameter data comprises at least one of: connection data or queue data.

20. The non-transitory machine-readable medium of claim 16 , wherein the resource deletion criterion comprises a threshold value and a confidence score.

Assignees

• Dell Products Lp

Inventors

• Prasad Parmeshwr
• Vishwakarma Rahul Deo

Classifications

• H04L67/562 Primary

  Brokering proxy services · CPC title

• H04L41/0622

  based on time · CPC title

• H04L67/51

  Discovery or management thereof, e.g. service location protocol [SLP] or web services · CPC title

• H04L67/60 Primary

  Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources (admission control or resource allocation H04L47/70) · CPC title