Linux & Docker · Lesson 3 of 7

Processes, services & networking

Checking CPU, memory and disk, managing processes and services, and basic network troubleshooting.

When an app is slow or down, these are the first commands to run:

top / htop — live CPU and memory by process.
ps aux | grep java — find a process and its PID; kill PID stops it.
df -h — disk space; free -h — memory.
systemctl status|start|restart nginx — manage services; journalctl -u nginx shows their logs.
ss -tulpn — which ports are open and by whom.
curl, ping, dig — test URLs, reachability and DNS.

Connect to remote servers with ssh user@server, using SSH keys rather than passwords.

Example in the terminal

Run these on your own computer

$ df -h /
Filesystem      Size  Used Avail Use% Mounted on
/dev/sda1        40G   38G  2.0G  95% /

$ free -h
               total   used   free
Mem:           7.8Gi  6.9Gi  0.9Gi

$ ps aux --sort=-%mem | head -3
USER   PID  %CPU %MEM COMMAND
app   1842  12.0 61.5 java -jar jobs-api.jar
root   901   0.3  2.1 nginx: master process

$ sudo systemctl restart jobs-api
$ systemctl is-active jobs-api
active

$ curl -s -o /dev/null -w "%{http_code}\n" http://localhost:8080/actuator/health
200

Lines starting with $ are commands you type; the rest is what the terminal prints. Use Linux, macOS, WSL on Windows or a free cloud shell. Names and paths are examples.

Key points

top, free -h and df -h show CPU, memory and disk.
ps finds processes; kill stops them.
systemctl manages services; journalctl shows their logs.
curl and ss check that an app is listening and responding.

Quick check

Disk shows 95% used and the app can't write logs. Which command confirmed the disk problem?

free -h df -h ping chmod

Practice task

Check your machine's disk usage, memory and the top 3 processes by memory, and write down which one uses the most.

Interview corner

A Linux server is slow. How do you troubleshoot?

Check load and the busiest processes with top, memory with free -h, disk space with df -h and disk I/O with iostat, read application and system logs, check network and dependencies with curl or ping, and compare against when the problem started (a deployment or traffic spike).

How do you see which process is using a port?

ss -tulpn (or netstat -tulpn, or lsof -i :8080) lists listening ports with the process name and PID.