Connectors

Observability: logs and metrics

Add Grafana, Loki, Prometheus, Datadog or other MCP tools so investigations can query real telemetry.

8 min read

Where to add tools

  1. Open Settings → Tools & MCP.
  2. Add the MCP server or connector for your stack (for example Grafana with Loki and Prometheus, or Datadog).
  3. Save credentials locally as the form describes. Test the connection.
  4. Open a ticket that mentions a service and time window. Ask Solvaro to pull error logs or the latency chart for that window.
Metrics chart inside Solvaro during an investigation
Live metrics chart from a connected observability source (demo data).

What good evidence looks like

  • Log lines with timestamps that match the ticket.
  • A metric spike aligned with the incident window.
  • A code citation that explains the log or metric.

If queries fail

Check the MCP tool status in Settings → Tools & MCP, confirm the time range, and retry. Investigate still works from the ticket and local code alone.