Ishikawa/Fishbone View
Fishbone diagrams are used in Container Observability to help users analyze the root cause of issues with predictive explanations.
To elaborate further with an example. If a service was in a different state at any observed time, knowing the current state, Fishbone analysis provides a hint on the expected next state. Such hints are very helpful to users to analyze what went wrong.
The primary predictive indicators (symptoms) of a problem help you to take any preventive actions. For example, if a service fails and the behavior model indicates higher CPU usage and more disk writes. You can check what caused more writes (and the associated increase in CPU usage).
Considering an anomaly occurred in the pod based on the alert rule, from the Fishbone analysis you can analyze the following details:
It represents all the possible issues with categories (dependencies) listed to investigate that led to permanent failure. With reference to the above figure, the pod is affected due to BackOff Restart, hence its categories are likely to be affected prior to complete failure.
Click the red color highlighted bone. A pop-up chart details the exact cause of an issue that triggered an alert as follows.
 |
The analysis helps you to take the appropriate actions to resolve the issue.
Related topics: