Preventative remediation of services

• US 10,261,849 B1
• Filed: 08/11/2017
• Issued: 04/16/2019
• Est. Priority Date: 08/11/2017
• Status: Active Grant

• Current Assignee: Electronic Arts Incorporated
• Original Assignee: Electronic Arts Incorporated
• Inventors: Burke, Maximilian Louis, Caballero, III, Oswaldo Carlos
• Primary Examiner(s): Bryan, Jason B
• Application Number:
  US15/675,624
• Time in Patent Office: 613 Days
• CPC Class Codes: A63F 13/35 Details of game servers
• CPC Class Codes: A63F 13/70 Game security or game management aspects
• CPC Class Codes: G06F 11/0709 in a distributed system consisting of a plurality of standalone computer nodes, e.g. clusters, client-server systems
• CPC Class Codes: G06F 11/0736 in functional embedded systems, i.e. in a data processing system designed as a combination of hardware and software dedicated to performing a certain function testing or monitoring of automated control systems G05B23/02
• CPC Class Codes: G06F 11/0754 by exceeding limits
• CPC Class Codes: G06F 11/0793 Remedial or corrective actions recovery from an exception in an instruction pipeline G06F9/3861; by retry G06F11/1402; for recovering from a failure of a protocol instance or entity H04L69/40
• CPC Class Codes: G06F 18/214 Generating training patterns; Bootstrap methods, e.g. bagging or boosting
• CPC Class Codes: G06N 3/04 Architecture, e.g. interconnection topology
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/084 Backpropagation, e.g. using gradient descent
• CPC Class Codes: G06V 10/82 using neural networks
• CPC Class Codes: G06V 10/94 Hardware or software architectures specially adapted for image or video understanding
• CPC Class Codes: G06V 20/62 Text, e.g. of license plates, overlay texts or captions on TV images
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/153 using recognition of characters or words
• CPC Class Codes: G06V 30/18057 Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]
• CPC Class Codes: G06V 30/19173 Classification techniques

Abstract

A remediation server utilizing a deep neural network to analyze a live service to predict when a service is about to experience a failure and determine the least intrusive method of remediation. The remediation server prioritizing maintaining the highest level of system availability when determining a remediation for a failed or failing service.

Claims

1. A computer-implemented method as implemented by a hardware processor of a remediation server comprising specific computer executable code for remediating a service failure, the method comprising:

receiving an operative log of a service that comprises operative status codes indicative of operative performance of the service;

scanning the operative log as an image for the operative status codes;

identifying a potential service failure of the service based at least in part on the scanned operative status codes of the operative log;

generating a status code image that is indicative of the operative status codes occurring in the service, wherein pixel bit values of the status code image indicate at least levels of occurrences of the operative status codes, and wherein the status code image is optimized for processing by a convolution neural network;

filtering, by way of the convolution neural network that uses one or more filtering parameters, the generated status code image, wherein the one or more filtering parameters are uniquely determined for the generated status code image;

determining a remediation for the potential service failure based at least in part on the filtered status code image that prioritizes an intrusivity of the determined remediation;

sending the determined remediation to the service to remediate the potential service failure; and

propagating the one or more filtering parameters and the determined remediation for the generated status code image to the remediation server.

2. The method of claim 1, wherein the operative log of the service is a text file containing the operative status codes in fixed locations, and wherein the operative status codes are conventional hypertext transfer protocol status codes.

3. The method of claim 2, wherein the remediation server is configured with the fixed locations of the operative status codes within the operative log to enable the remediation server to omit scanning of a remainder of the operative log.

4. The method of claim 3, wherein enabling the remediation server to omit scanning of the remainder of the operative log contributes to a reduction of time taken to remediate the service.

5. The method of claim 3, wherein the remediation server is further configured to scan the remainder of the operative log for additional operative metrics of the service.

6. The method of claim 2, wherein the scanning of the operative log as the image is performed character by character using character text filters, and wherein the character text filters comprise a pixel size equivalent to a scanned status code character, such that the remediation server performs a one-to-one pixel comparison for identifying the potential service failure.

7. The method of claim 1, wherein a current service failure is further identified when scanning the operative log.

8. The method of claim 1, wherein the convolutional neural network utilizes remediation filters to filter the generated status code image, and wherein the remediation filters are image based convolution kernels with a pixel bit value equal to that of the generated status code image.

9. The method of claim 8, wherein the remediation filters include a scalar value equivalent to a reciprocal value of a maximum bit value of the generated status code, and wherein inclusion of the scalar value in the remediation filters enables the convolution neural network to omit applying weights to inputs of the convolution neural network.

10. The method of claim 9, wherein enabling the convolution neural network to omit applying the weights to the inputs of the convolution neural network contributes to a reduction of time taken to remediate the service.

11. A remediation server system comprising;

one or more processors; and

a memory comprising instructions stored thereon that, when executed by the one or more processors, cause the one or more processors to;

receive an operative log of a service that comprises operative status codes indicative of operative performance of the service;

scan the operative log for the operative status codes;

identify a potential service failure of the service based at least in part on the scanned operative status codes of the operative log;

generate a status code image that is indicative of the operative status codes occurring in the service, wherein pixel bit values of the status code image indicate at least levels of occurrences of the operative status codes, and wherein the status code image is optimized for processing by a convolution neural network;

filter, by way of the convolution neural network that uses one or more filtering parameters, the generated status code image, wherein the one or more filtering parameters are uniquely determined for the generated status code image;

determine a remediation for the potential service failure based at least in part on the filtered status code image that prioritizes an intrusivity of the determined remediation;

send the remediation to the service to remediate the potential service failure; and

propagate the one or more filtering parameters and the determined remediation for the generated status code image to the remediation server.

12. The system of claim 11, wherein the operative log of the service is a text file containing the operative service codes in fixed locations and wherein the operative status codes are conventional hypertext transfer protocol status codes.

13. The system of claim 12, wherein the remediation server is configured with the fixed locations of the operative status codes within the operative log to enable the remediation server to omit scanning of a remainder of the operative log.

14. The system of claim 13, wherein the remediation server is further configured to scan the remainder of the operative log for additional operative metrics of the service.

15. The system of claim 13, wherein enabling the remediation server to omit the scanning of the remainder of the operative log contributes to a reduction of time taken to remediate the service.

16. The system of claim 12, wherein the scan of the operative log is performed character by character using character text filters, and wherein the character text filters comprise a pixel size equivalent to a scanned status code character, such that the remediation server performs a one-to-one pixel comparison for identifying the potential service failure.

17. The system of claim 11, wherein a current service failure is further identified when scanning the operative log.

18. The system of claim 11, wherein the convolutional neural network utilizes remediation filters to filter the generated status code image, and wherein the remediation filters are image based convolution kernels with a pixel bit value equal to that of the generated status code image.

19. The system of claim 18, wherein the remediation filters include a scalar value equivalent to a reciprocal value of a maximum bit value of the generated status code, and wherein inclusion of the scalar value in the remediation filters enables the convolution neural network to omit applying weights to inputs of the convolution neural network.

20. The system of claim 19, wherein enabling the convolution neural network to omit applying the weights to the inputs of the convolution neural network contributes to a reduction of time taken to remediate the service.