loading page

Root Cause Analysis Bot using Machine Learning Techniques
  • Tapan Behera ,
  • Kumud Tripathi
Tapan Behera

Corresponding Author:[email protected]

Author Profile
Kumud Tripathi
Author Profile

Abstract

In this world of quick delivery of quality products, DevOps test automation plays a vital role. This triggers the automation build to test whether the product quality is a good fit for the release or not? When multiple test cases of a test suite got failed, it takes lots of time for the developer and analyst to analyze the error logs of each test case and do the root cause analysis (RCA). Second, Enterprises with complex distributed applications designed with multi-tier architecture systems and with a lot of third-party applications communicating with each other, cannot prevent failures from happening. As a result, finding the cause of failures becomes a real challenge, as it requires cross-team cooperation, log analysis, and many more. As a result, a lot of investigation was required in order to perform a Root Cause Analysis and becomes part of the daily routine, hindering the organization’s productivity. To solve this time-consuming process, we are introducing the Root Cause Analysis Bot (RCA Bot) in this paper. This bot will analyze the failure logs, error messages, errorClasses, descriptions, and error codes and apply machine learning (ML) techniques to predict the root cause of the failure with a percentage of the prediction accuracy.