VDS
VDS 24/7 uptime and monitoring tips
Tabi Cloud runs servers with 24/7 uptime, but your own software also has to stay up. Services crash, disks fill and certificates expire. Good monitoring catches those before users do. Commands target Ubuntu and Debian and may vary elsewhere.
1. Auto-restart your services
Use systemd so a crashed program comes back.
- Add Restart=always and RestartSec=5 to the unit file
- sudo systemctl enable NAME starts it at boot
- systemctl status NAME shows its state
2. Watch resources
- htop, free -h and df -h for a quick look
- uptime for load
- journalctl -p err -b lists errors since boot
3. Alert on the important things
- External checkers such as UptimeRobot can ping your site or port from outside
- Send alerts to email or a Discord webhook
- Monitor disk space above 80 percent, since it is the most common silent killer
4. A tiny disk alert script
- Create check-disk.sh containing: USE=$(df / --output=pcent | tail -1 | tr -dc '0-9')
- Add: if [ "$USE" -gt 85 ]; then echo "Disk high" | mail -s "VDS alert" [email protected]; fi
- chmod +x check-disk.sh
- Schedule it with crontab -e, for example */15 * * * * /home/deploy/check-disk.sh
Mail needs a mail tool installed, so you can instead post to a Discord webhook with curl.
5. Keep up with updates
- sudo apt update and sudo apt upgrade on a schedule
- unattended-upgrades for security patches
- Reboot in quiet hours when a kernel update needs it
6. Certificates and domains
- sudo certbot renew --dry-run tests renewals
- Note your domain renewal date
7. Have a recovery plan
- Keep backups off the server, using rsync or tar
- Write down how to rebuild: packages, configs and ports
- Know how to reach support
Reaching us
If something looks wrong on the platform side, open a ticket on our Discord or email support. Include your server IP, the time the issue started and what you see in the logs.
Tips
- Test your alerts so you know they work
- Check logs weekly
- Review usage monthly and change your plan if needed