Investigations are automated analysis workflows that help identify and diagnose issues within Kubernetes clusters. They reduce Mean Time To Repair (MTTR) by systematically analyzing cluster state and suggesting remediation steps.

Currently, Investigations can be initiated manually through the Agentkube dashboard. Automatic triggering of response protocol will be added in future releases.

Agentkube’s Investigation feature enables quick identification of cluster issues through automated analysis tools. It employs a systematic approach which is defined by the you in response protocol, it examines cluster state ensure comprehensive problem diagnosis.

​
Investigation Process

​
1. Issue Detection

  • Identifies anomalies in cluster behavior
  • Gathers initial diagnostic information
  • Determines affected components

​
2. Data Collection

  • Collects relevant logs
  • Gathers metrics
  • Captures cluster state

​
3. Findings

  • Identifies potential root causes
  • Provides evidence for conclusions
  • Suggests remediation steps
  • Assesses impact
Response ProtocolAI Editor