Monitoring Non-SSH or Cloud-API-Mode Servers
Who this is for
Users who have imported existing servers that do not have the monitoring agent installed, or who manage servers where SSH access is restricted.
What you will complete
Understand the different monitoring modes, what data is available for each mode, and how to enable deeper monitoring on imported servers.
Before you begin
- You need at least one server in your organization.
- Admin role required to change server monitoring settings.
The two monitoring modes
Full monitoring (agent-based)
When a server is provisioned by CloudAIPilot, the monitoring agent is installed automatically. This provides:
- Real-time CPU, RAM, disk, and network metrics every 60 seconds
- Load average tracking
- Service status visibility
- SSH-accessible logs (when File and Log Access is enabled for AI Pilot)
- Full alert rule support for all metric types
Server card appearance: Shows live metric values with colored status indicator.
Limited monitoring (cloud API mode)
When a server is imported without the agent, CloudAIPilot can still fetch some information using your cloud provider's API. This provides:
- Server power state (running, stopped)
- Basic cloud-level metrics (if the provider exposes them — not always available)
- Provider-reported status
Server card appearance: Shows a grey or partial status indicator. Metric graphs may show "No data" or limited data.
How to tell which mode a server is in
- Go to Servers and click on the server.
- On the Overview tab, look for a monitoring status indicator or a message like "Monitoring agent not detected."
- If the monitoring dashboard shows "No data" for this server, the agent is not reporting.
Step-by-step: install the monitoring agent on an imported server
If you imported a server and want full monitoring, you need to install the agent manually.
- Go to Servers and click on the imported server.
- On the Overview tab, look for the Install Agent or Enable Full Monitoring button.
- CloudAIPilot will display the installation command — a one-line script to run on the server.
- SSH into your server as root or a user with sudo privileges.
- Paste and run the installation command.
- Wait 60–90 seconds for the first metric report.
- Return to the Monitoring dashboard — the server card should now show live metrics.
If you do not see an Install Agent button: The agent may already be installed but experiencing a connectivity issue. Check that the server can reach the CloudAIPilot platform endpoint on outbound HTTPS (port 443).
Alert rules on non-agent servers
Alert rules for CPU, RAM, disk, and load average require the monitoring agent. These rules will show as "No data" on imported servers without the agent.
The only alerts available on cloud-API-mode servers:
- Server power state change (if the provider reports it)
- Connection health (if the server becomes unreachable)
What success looks like
- After installing the agent, the server card in the Monitoring dashboard shows live CPU, RAM, and disk values within 2 minutes.
- Alert rules begin evaluating and may fire if thresholds are immediately crossed.
- The server status indicator turns green.
Common errors and fixes
"Installation command failed with 'Permission denied'" Cause: The command was run as a non-root user without sudo. Fix: Prepend sudo to the installation command, or switch to the root user before running it.
"Agent installed but still showing No data after 5 minutes" Cause: The server may have an outbound firewall rule blocking HTTPS traffic to the CloudAIPilot platform. Fix: Verify outbound port 443 is open from the server. Check the firewall rules in your cloud provider's console.
"The Install Agent button is not visible on my server" Cause: The server may be in a stopped state — the agent requires a running server to install. Fix: Start the server first, then retry.