Platforms

How do you monitor a RunCloud server?

Monitor a RunCloud server: the five commands that show its health, then the Opservo agent over SSH in 30 seconds for alerts, causes and fixes.

By , Director of Softuvo Solutions (LinkedIn) · Updated

In short

To monitor a RunCloud server, watch what RunCloud does not explain: services that stop, a disk filling, memory running out, exposed ports and failed logins. Check them by hand with five commands, or install the Opservo agent over SSH (RunCloud has no documented way to run your own script as root): one command, about 30 seconds, nothing changes in RunCloud, no inbound port.

What do you need?

  • •A RunCloud server (RunCloud supports Ubuntu LTS, which runs systemd)
  • •Root SSH access. If you closed port 22 after connecting the server to RunCloud, open it for your own IP while you install
  • •An Opservo workspace; Free covers up to 2 servers
  • •Outbound HTTPS (port 443) to api.getopservo.com

How do you install it?

In Opservo, click Add a new server, then Generate install command. The token in it works once and expires after one hour. SSH to the server and run it as root:

curl -sSL https://api.getopservo.com/install.sh | sudo env OPSERVO_API_URL=https://api.getopservo.com OPSERVO_TOKEN=<token> bash

Deploying firewall rules from the RunCloud dashboard overwrites rules added by hand. The agent needs no inbound rule, so this does not affect it.

What should you check on a RunCloud server by hand?

CheckCommandProblem when
Stopped servicessystemctl --failedAny unit is listed
Web, PHP and database servicessystemctl list-units --type=service --state=running | grep -Ei "nginx|apache|php|mysql|maria|redis"One you expect is missing
Diskdf -h; df -iA mount over 90 percent, or inodes full
Memoryfree -m; dmesg -T | grep -i "killed process"Available near zero, or OOM kills
Exposed databasesss -tlnp | grep -E ":3306|:5432|:6379"Bound to 0.0.0.0 rather than 127.0.0.1

A five-minute health check for a RunCloud server

The free one-command health check runs all of these and ten more, read-only, and prints the fix for each. The agent runs them every minute and alerts you.

What do you see afterwards?

Each server gets a 0–100 health score and findings with the command that fixes them: CPU, memory, disk and load; systemd services such as PHP-FPM, MySQL or MariaDB and Redis; listening ports; security checks such as exposed database ports and failed SSH logins; and kernel-log errors. The agent reads domains and SSL expiry from the standard nginx and Apache config folders. If your RunCloud sites do not appear on the Domains tab, add an uptime monitor for each site; it checks the site from outside and tracks its certificate.

Last updated 2026-10-28.

FAQ

Related questions

Does it change anything in RunCloud?+

No. The agent is a separate systemd service. It is read-only by default and never reads .env files.

Can Opservo run fixes on the server?+

Only if an owner or admin turns on remote actions for that server in the dashboard, on a paid plan; it is off by default. Every command is confirmed and audited.

How do you remove it?+

Run curl -sSL https://api.getopservo.com/install.sh | sudo env OPSERVO_UNINSTALL=1 bash, then remove or archive the server in the fleet table.