Failure Root Cause Analysis for Microservices, Explained

Determining the root causes of observed failures is a main issue in microservice-based applications. Unfortunately, available root cause analysis techniques do not focus on explaining how root failures actually caused the observed failure. On the other hand, the availability of such explanations would greatly help to pick adequate countermeasures, e.g., by introducing circuit breakers or bulkheads. We hence present a declarative root cause analysis technique, which can determine the cascading failures that possibly caused an observed failure, identifying also (or starting from) a root cause. We also introduce a prototype implementation of our technique, and we use it to assess our technique by means of controlled experiments.