RunCloud AI Monitoring helps you detect and understand server issues faster, without changing anything on your server.
What is RunCloud AI Monitoring?
RunCloud AI Monitoring is an intelligent monitoring system that analyzes your server in real time and alerts you when something needs attention. Instead of just notifying you that a metric has spiked, it acts as your server assistant to help you understand:
- What’s happening
- Why it might be happening
- What impact could it have
- What you can do next
We understand that introducing AI to your critical infrastructure can be disastrous if not done properly. That is why RunCloud AI Monitoring is entirely optional, and if you choose to enable it, it operates under a strict “look, but don’t touch” philosophy.
The AI acts as an advisor, not an administrator. It is designed to be highly secure, predictable, and transparent, ensuring your server remains strictly under your control:
- It Is Read-Only: The AI Agent only observes your server.
- It Does Not Execute Commands: It will never automatically kill processes, delete files, or modify server configurations.
- Manual Execution Required: All suggested remediation steps and CLI commands must be reviewed and manually executed by you via SSH or your terminal.
- Verify Before Acting: AI-generated insights are designed to be helpful, but they are not perfect. Recommendations may be incomplete, and you should always verify the root cause before executing commands on a production server.
This architecture ensures that you receive intelligent, actionable remediation steps while retaining absolute authority over your production environment.
How it Works
RunCloud AI Monitoring continuously monitors your server and detects unusual behavior. Here’s what happens behind the scenes:
- Monitoring: RunCloud AI Monitoring tracks key system signals, including CPU, memory, and disk usage, server load, service status, and system logs.
- Detection: It identifies anomalies when behavior goes beyond safe thresholds (e.g., disk usage > 90%, load average > 5.00) or deviates from normal patterns.
- Analysis: When an issue is detected, the AI agent analyzes the available data to determine possible causes, affected components, and severities (‘Info’, ‘Warning’, ‘Critical’).
- Recommendations: You receive a detailed pop-up containing actionable remediation steps (complete with CLI commands to verify and fix the issue, such as df -h or systemctl status fail2ban).
How to Enable AI Monitoring in RunCloud
To enable RunCloud AI Monitoring:
- Go to your server dashboard in the RunCloud management panel.
- Click on the Install RunCloud AI button in the top-right corner of the screen.
- A consent notice will open in a new window. Accept the terms and conditions, then click “Enable AI Monitoring”.
Once enabled, the AI Agent is installed on your server, monitoring becomes active, and you’ll start receiving alerts when issues are detected.
Viewing Insights & The Analysis Dashboard
Once an issue is detected, the AI Agent logs it in your dashboard so you can investigate it further.
Accessing the AI Analysis Page
- Navigate to your server’s main dashboard.
- Open the existing installation/monitoring dropdown menu.
- Click the View AI Analysis link (see screenshot below).
Understanding the AI Analysis Page
The AI Analysis page contains a centralized log of all anomalies on your server.
- Severity Levels: Alerts are categorized by severity: Info, Warning, and Critical.
- Table Summary: You will see a table displaying the Diagnosis, Severity, and Date (all generated dynamically by the AI Agent).
- Search: Use the built-in search box to filter past alerts.
Detailed Analysis Popup
You can click the View Analysis button next to any AI log to open a detailed pop-up. This pop-up includes:
- Title & Severity Badge: Generated by the AI.
- Root Cause: Explains why the issue is occurring (e.g., “The configuration file is improperly formatted”).
- Potential Impact: Explains the risk (e.g., “If this continues, the server may struggle to handle requests”).
- Remediation Steps: Step-by-step guidance. The AI provides actual terminal commands (such as journalctl -u fail2ban -n 50 or top -b -n 1 | head -n 20) to help you troubleshoot.
- Next Steps: Suggested long-term actions (e.g., clearing disk space, modifying Cloudflare settings).
Troubleshooting & Known Issues
Why am I receiving standard resource alerts, but no AI Monitoring insights?
If you are receiving standard RunCloud resource alerts (such as email warnings for high CPU load) but the AI Monitoring system is not sending notifications, there may be a TLS connection issue preventing the agent from communicating with the system. To fix this, simply uninstall and reinstall the AI Agent on your server. This will refresh the TLS checks and restore normal communication.
Why am I not receiving any notifications at all?
If you aren’t getting any alerts at all, your notification channels likely haven’t been configured yet. To receive alerts from the AI Agent, you must enable and set up your preferred channels (such as Email or Slack) in your RunCloud dashboard. For step-by-step instructions, please visit our guide on Setting Up Notification Channels.
Why did I get a “monthly usage limit” notification, and why is my AI paused?
RunCloud AI Monitoring operates on a monthly analysis quota based on your current billing cycle. If your server experiences an unusually high number of anomalies in a short period, the AI Agent may reach this monthly limit. When this happens, you will receive a message stating: “Your AI Agent has reached its monthly usage limit.”
Your AI-powered features will be temporarily paused until your next billing cycle begins, at which point they will resume automatically. Please note that your standard RunCloud server monitoring will continue to function normally even while the AI is paused.