Servers

How to use AI server diagnosis

~2 min read

DeployFly's AI Diagnosis feature reads your server's logs, service statuses, and resource metrics, then provides a plain-English summary of detected issues with suggested remediation steps.

Steps

  1. 1

    Open Diagnosis

    Open your server and click "Diagnosis" in the sidebar.

  2. 2

    Start a diagnosis

    Click "Run Diagnosis". DeployFly collects data from your server over SSH — this takes 10–20 seconds. You will see a progress indicator while it works.

  3. 3

    Read the report

    The diagnosis report appears in sections: Resource Health (CPU, RAM, Disk), Service Status (which services are running or failed), Log Analysis (recent errors grouped by severity), and Recommendations (specific actions to fix detected issues).

  4. 4

    Apply a recommendation

    Each recommendation shows a "Fix" button where applicable. Clicking it runs the suggested command on your server and shows the output. Review the output carefully before applying.

  5. 5

    Re-run after fixing

    After applying fixes, click "Run Diagnosis" again to verify the issues are resolved. The report refreshes with current server state.

Watch out

  • Applying fixes runs real commands on your server. Only use "Fix" buttons for issues you understand.
  • Diagnosis collects log data — avoid running it on servers with very large log files (>500 MB) as it may slow temporarily.

Tips

  • Run a diagnosis whenever a site goes down or becomes slow — it often pinpoints the root cause in seconds.
  • The AI analysis is a starting point, not a definitive fix. Always read the full recommendation before clicking "Fix".
  • Diagnosis history is saved. Use the History tab to compare the current report against previous ones.

Related guides