Storage, Disks and Filesystems

What is using space?

df -h                                   # per filesystem
sudo du -xh / --max-depth=1 | sort -h   # biggest top-level folders
sudo du -sh /var/log/* | sort -h | tail
sudo find / -xdev -type f -size +500M -exec ls -lh {} \;

Usual culprits: logs, old releases, Docker images (docker system df), database files, package caches (sudo apt clean).

A file deleted while a process still has it open keeps using space. Find it with sudo lsof +L1, then restart that process.

Inodes

df -i shows inodes, the slots that track files. Millions of tiny files (sessions, caches) can exhaust inodes while df -h still shows free space, and writes fail with "No space left on device".

Adding a disk

lsblk                                  # find the new disk, e.g. /dev/sdb
sudo mkfs.ext4 /dev/sdb                # format it (destroys data on it!)
sudo mkdir -p /mnt/data
sudo blkid /dev/sdb                    # get its UUID

Add it to /etc/fstab so it mounts at boot, by UUID (device names can change):

UUID=3f1c...  /mnt/data  ext4  defaults,nofail  0  2

sudo mount -a          # mount everything in fstab: test BEFORE rebooting
df -h /mnt/data

nofail lets the server boot even if the disk is missing. A broken fstab line without it can leave a cloud server unbootable.

Swap

Swap is disk used as overflow memory. On small servers (512 MB to 2 GB) a little swap prevents out-of-memory crashes during spikes:

sudo fallocate -l 1G /swapfile
sudo chmod 600 /swapfile
sudo mkswap /swapfile && sudo swapon /swapfile
echo '/swapfile none swap sw 0 0' | sudo tee -a /etc/fstab
sudo sysctl vm.swappiness=10

Swap is a safety net, not extra RAM. If a server constantly swaps, it needs more memory or fewer workers.

Check yourself

• df -h shows free space but writes fail. What do you check?
• Why mount by UUID?
• What does nofail protect against?