open
  1 (866) 866-2320 Resources Events Blog

How to Use Anomaly and Causality Analysis to Determine Root-Cause

Blog

How to Use Anomaly and Causality Analysis to Determine Root-Cause

About

This content is brought to you by Evolven. Evolven Change Analytics is a unique AIOps solution that tracks and analyzes all actual changes carried out in the enterprise cloud environment. Evolven helps leading enterprises cut the number of incidents, slash troubleshoot time, and eliminate unauthorized changes. Learn more

Today, enterprises are recognizing the importance of providing uninterrupted digital services as customers rely on them for their daily work and living.

However, technology is never infallible, especially as IT environments today are becoming more and more complex and agile, with millions of changes being implemented on a daily basis across the enterprise cloud.

As every performance issue impacting the end-user experience can cost enterprises huge losses in revenue, competitiveness and brand reputation, finding the root-cause of issues to rapidly recover the issues, and also prevent them from reoccurring, has become critical.

Today, Anomaly & Causality Analysis is a critical technique to determining root-cause.

Detecting a Problem: Anomaly Analysis

Anomaly analysis works in parallel with suspicious behavior detection to determine when a system is behaving differently, helping ITOps discover a performance issue that needs to be addressed.

After gathering a historical baseline of every single metric across an application, Anomaly analysis learns how a system typically behaves so it can flag any differences that do not appear as typical behavior.

Automated anomaly detection and alerting on anomalies allow accurate early detection of symptoms indicating a brewing issue.

Furthermore, this approach eliminates predetermined alerting rules that frequently trigger alerts when there is no issue, creating alert storm and noise, or trigger alerts too late when the issue has already been noticed by the users.

Automated anomaly detection and alerting on anomalies allow accurate early detection of symptoms indicating a brewing issue.

Determining Why the Problem Occured: Causality Analysis

Once an anomaly is detected, ITOps needs to understand what the sequence of events were that propagated and resulted in the anomaly and performance issue.

According to Gartner, 85% of performance issues can be traced back to changes.

So the next question is, what changed?

Causality Analysis helps ITOps answer this question by answering:

  1. Which changes are affecting the environment? Such as changes in configuration, deployed application changes or changes in data processed by the application
  2. Which changes are relevant to the incident?
  3. What is the temporal relationship between the changes and the incident?
  4. How has the change effect propagated across environment topology?
  5. What changes actively resulted in the anomalies, or differences in the behavior system?

“Causality analysis is the key technique in effective root cause analysis” - Bostjan Kaluza, Chief Data Scientist of Evolven.

Closing the Loop to Prevent Performance Issues

Once causality analysis reveals the root-cause, ITOps, SRE or DevOps can create a new rule for the system’s historical baseline and suspicious behavior modeling to prevent similar incidents from reoccurring in the future.

Using integrated technologies of Evolven and AppDynamics, such rules can be generated automatically based on the results of causal analysis.

Smarter Together

With the Evolven and AppDynamics® Integrated Solution, enterprises will have anomaly detection through AppDynamics, and causality analysis through Evolven - the best of both worlds.

Furthermore, while AppDynamics tells customers the health of applications, answering the question “What’s Wrong?”, when and where an issue exactly started, and what its impact is, Evolven will answer "What Changed?”, focusing on actual granular changes it has detected that are related to the particular issue or anomaly that needs to be investigated.

Enterprises using the Evolven and AppDynamics® Integrated Solution instantly had 50% faster incident resolution by using automated anomaly and causality analysis to find the root-cause.

See integrated Evolven and AppDynamics solution in action - contact us now to schedule a demo.

About the Author
Nicole Prybula

Field Marketing Manager