Basic Architecture
The environement (Kubernetes, Applications, …) create logs that are being fed into a Log Management Tool (ELK in this case).
- External Systems generate Alerts and send them into the CP4AIOPS for Event Grouping.
- At the same time CP4AIOPS ingests the raw logs coming from the Log Management Tool (ELK) and looks for anomalies in the stream based on the trained model.
- It also ingests Metric Data and looks for anomalies
- If it finds an anomaly (logs and/or metrics) it forwards it to the Event Grouping as well.
- Out of this, CP4AIOPS creates a Story that is being enriched with Topology (Localization and Blast Radius) and with Similar Incidents that might help correct the problem.
- The Story is then sent to Slack.
- A Runbook is available to correct the problem but not launched automatically.