Archives, disks and network tools: tar, zip, df, du, mount, curl, wget, ip and ss

Three everyday admin jobs fill this unit: packing files (backups, uploads, downloads come as .tar.gz or .zip), managing disk space (a full disk is one of the most common reasons websites and databases suddenly stop working), and checking the network (is the server online, which ports are listening, can it reach an API?). These commands are used by developers, sysadmins, network engineers and support staff daily.

Archives and compression

An archive bundles many files into one; compression makes it smaller. On Linux the classic is tar (tape archive) plus gzip.

tar -czf site-backup.tar.gz /var/www/site     # create (c), gzip (z), file name (f)
tar -tzf site-backup.tar.gz                   # list contents (t)
tar -xzf site-backup.tar.gz                   # extract (x) here
tar -xzf site-backup.tar.gz -C /tmp/restore   # extract into a folder
tar -cJf logs.tar.xz /var/log/nginx           # xz: smaller, slower

Remember the letters: create, extract, list, z gzip, J xz, v verbose, f file.

sudo apt install zip unzip
zip -r photos.zip photos/
unzip photos.zip -d extracted/
unzip -l photos.zip              # list contents
gzip big.log                     # becomes big.log.gz
zcat big.log.gz | less           # read without extracting

Disk space

df -h                    # free space per file system (human-readable)
df -i                    # inodes: running out of these also stops new files
du -sh /var/www          # size of one folder
du -h --max-depth=1 /var | sort -h      # what's big inside /var
sudo du -xh / --max-depth=2 2>/dev/null | sort -h | tail -20   # biggest folders on the system disk

Interactive and friendlier: sudo apt install ncdu, then sudo ncdu /.

Common space eaters on servers:

Disks, partitions and mounting

lsblk                     # disks and partitions as a tree (sda, sdb, nvme0n1...)
sudo fdisk -l             # detailed partition info
mount | column -t         # what's mounted where

Mount a USB drive or extra disk:

sudo mkdir -p /mnt/usb
sudo mount /dev/sdb1 /mnt/usb
ls /mnt/usb
sudo umount /mnt/usb      # always unmount before removing

To mount a disk automatically at boot, add it to /etc/fstab (using its UUID from sudo blkid). A mistake in fstab can stop a server from booting, so test with sudo mount -a before rebooting.

Formatting erases data: sudo mkfs.ext4 /dev/sdb1 (double-check the device name with lsblk first!).

curl and wget: talking to the web

wget https://example.com/file.zip              # download a file
wget -c https://example.com/big.iso            # continue an interrupted download
curl https://example.com                       # print a page
curl -o page.html https://example.com          # save to a file
curl -I https://example.com                    # headers only: status code, server, redirects
curl -L http://example.com                     # follow redirects
curl -s https://api.github.com/repos/torvalds/linux | head -20    # call an API
curl -X POST -H "Content-Type: application/json" \
     -d '{"name":"Amina"}' https://httpbin.org/post               # send JSON
curl -w "%{http_code} %{time_total}s\n" -o /dev/null -s https://example.com   # status and timing

curl -I is a quick way to check if a site is up and whether HTTPS redirects work:

HTTP/2 200
server: nginx
content-type: text/html

Network information

ip a                      # IP addresses of each interface (ip addr)
ip r                      # routing table, including the default gateway
ping -c 4 8.8.8.8         # can we reach the internet by IP?
ping -c 4 google.com      # does DNS work too?
dig example.com +short    # DNS lookup (sudo apt install dnsutils)
traceroute example.com    # path packets take (sudo apt install traceroute)
ss -tulpn                 # listening ports and which program owns them (sudo for all names)
hostname -I               # this machine's IP addresses
cat /etc/resolv.conf      # DNS servers in use

Reading ss -tulpn:

tcp  LISTEN 0 511  0.0.0.0:80    0.0.0.0:*  users:(("nginx",pid=812))
tcp  LISTEN 0 151  127.0.0.1:3306 0.0.0.0:* users:(("mysqld",pid=640))

Nginx listens on port 80 on all addresses (0.0.0.0): reachable from outside. MySQL listens only on 127.0.0.1 (localhost): good, it isn't exposed to the internet.

The networking subject explains IP addresses, DNS and ports in depth.

Troubleshooting routines

"The website stopped working, and errors say no space left on device"

1. df -h: which file system is at 100%?
2. sudo du -xh / --max-depth=2 | sort -h | tail: find the biggest folders.
3. Clean safely: rotate/compress logs, delete old backups, apt clean.
4. Restart affected services (sudo systemctl restart mysql nginx php8.3-fpm, adjusting names to your versions).
5. Prevent it: logrotate, backup retention, a disk-usage alert script (see bash scripting).

"The site is unreachable"

1. From your computer: ping the domain and curl -I https://domain.
2. DNS: dig domain +short: does it return the server's IP?
3. On the server: systemctl status nginx: is the web server running?
4. ss -tulpn | grep -E ':80|:443': is it listening?
5. Firewall: sudo ufw status: are 80/443 allowed?
6. Logs: sudo tail -50 /var/log/nginx/error.log.

Summary

• tar -czf creates and tar -xzf extracts .tar.gz; zip/unzip for Windows users; test restores.
• df -h shows free space; du -sh and ncdu find big folders; logs and backups are common culprits.
• lsblk, mount, umount and /etc/fstab manage disks; formatting erases data.
• curl and wget download and test web requests (curl -I for status/headers).
• ip a, ip r, ping, dig, traceroute and ss -tulpn check network setup and listening ports.