Troubleshooting, Health Checks, and Support Diagnostics
This domain tests troubleshooting, health checks, and support diagnostics: how to prove an agent is healthy, isolate connectivity from configuration, and collect evidence for support.
Health check sequence
Check agent service status, then the agent log, then outbound connectivity to the tenant endpoint on 443, DNS, proxy, clock skew, and disk space. In Kubernetes, check the operator and DaemonSet pod logs, events, and the custom resource status. Confirm the agent version and key match the tenant.
Diagnostics
Use the agent status command and support diagnostic bundle, plus UI agent status and self-hosted backend health endpoints. Capture timestamps, host IDs, agent version, endpoint, and exact error text. Reproduce with a minimal configuration before changing many settings.
Worked example
An agent is installed but never appears. systemctl status instana-agent shows running; the log shows TLS handshake timeout. curl -v https://<tenant-endpoint>:443 also times out, while DNS resolves. The firewall team confirms outbound 443 is blocked. Opening the destination resolves the issue; reinstalling the agent would not have helped.
Traps
Restarting repeatedly without reading logs, testing only inbound rules, using the wrong regional endpoint, ignoring proxy environment variables, and collecting a support bundle after deleting logs are common errors. Change one variable at a time and record results.