Machine Learning (ML) alerts
Container Observability creates an alert whenever there is a deviation of the container from the normal behavior and ML learns such deviations. This behavior model is built on a curated set of container metrics. Container Observability can trigger ML-based alerts for containers, nodes, and App Objects.
Note
Note: Currently, you cannot define a new set of conditions to create a new ML model as Container Observability does not support it in the system.
ML-based Alerts help you:
Compare the current metric value with what the ML model expects and trigger an alert if the behavior is not as expected by the model.
Track the occurrence time and updated time of the anomaly.
Allows you to view the RCA of the anomaly that occurred to determine the possible cause and its dependencies that can likely cause an error. For more information, see Causal analysis for ML-based Anomaly.
Enables predictive analysis to isolate the cause of the problem.
Related topics: