Chapter 4: Linux Ops Refresh for DevOps
4.5 Disk, process, ports — the triage trio
Disk — df -h
df -h
What you see: size, used, available, use% per mount. If / is 100%, new deploys, logs and Docker layers fail in confusing ways. Clear old logs or unused images (carefully) — do not delete random system folders.
Process — ps
ps aux | head
ps aux | grep -i nginx
What you see: process list with user and command. Confirm the app you expect is alive.
Ports — ss (logical counters)
ss -tlnp
What you see: TCP ports in listen mode, and often the process name. Remember: one building (IP), many published ports (ports). Port 22 is SSH. Port 80 is HTTP. Your app might use 3000 or 8080 on the host during labs — still a logical counter, not hardware.
Ravindra Bagale's Tip
Khup students localhost var kaam karta aani Security Group madhe port open nahi karta — "AWS down" mhanun blame. Aadhi server-local curl, mag SG / firewall. Farak lakshat theva!
Steps — five-minute triage
df -h— any mount ≥ 90%?systemctl status <service>— active?ss -tlnp— expected port listening?curl -I http://localhost:<port>— what status?journalctl -u <service> -n 50— last error line?