Linux Disk Management: Partition, Format, Mount and Check Disks

lsblk -f                                      # find the new disk, e.g. /dev/sdb
sudo parted /dev/sdb --script mklabel gpt mkpart data ext4 0% 100%
sudo mkfs.ext4 -L data /dev/sdb1
sudo mkdir -p /data
echo "UUID=$(sudo blkid -s UUID -o value /dev/sdb1) /data ext4 defaults,nofail 0 2" | sudo tee -a /etc/fstab
sudo systemctl daemon-reload
sudo mount -a && df -h /data

1. See your disks and partitions

# tree of disks, partitions, filesystems and mount points
lsblk -f
# sizes and models
lsblk -o NAME,SIZE,TYPE,MODEL,SERIAL,TRAN
# UUIDs and filesystem types
sudo blkid
# usage of mounted filesystems
df -hT
# partition tables in detail
sudo fdisk -l

Example output of lsblk -f:

NAME        FSTYPE FSVER LABEL UUID                                 FSAVAIL FSUSE% MOUNTPOINTS
sda
└─sda1      ext4   1.0   data  3f6c2a1e-9b0d-4e5f-8a7b-1c2d3e4f5a6b  850.2G     4% /data
nvme0n1
├─nvme0n1p1 vfat   FAT32       A1B2-C3D4                               1G     2% /boot/efi
└─nvme0n1p2 ext4   1.0         8d7e6f5a-4b3c-2d1e-0f9a-8b7c6d5e4f3a  120.5G    48% /

Naming: SATA/USB disks are /dev/sda, /dev/sdb… with partitions sda1; NVMe disks are /dev/nvme0n1 with partitions nvme0n1p1. Names can change between boots, which is why fstab should use UUIDs. To find what's filling a disk, use the find largest files script or sudo du -xh / --max-depth=1 | sort -h.

2. Partition a disk

Use GPT for every new disk unless an old BIOS machine needs to boot from it with MBR. Three tools:

• fdisk (interactive, GPT and MBR): sudo fdisk /dev/sdb, then g new GPT table, n new partition, t type, p print, w write. Nothing changes until w.
• parted (scriptable, good for automation):

  sudo parted /dev/sdb --script \
    mklabel gpt \
    mkpart backup ext4 0% 50% \
    mkpart media ext4 50% 100%
  sudo parted /dev/sdb print

• cfdisk or GNOME Disks / KDE Partition Manager / GParted for a visual interface.

Using percentages or MiB boundaries in parted keeps partitions aligned for SSD performance. After changing a disk that is in use, run sudo partprobe /dev/sdb so the kernel rereads the table.

3. Create a filesystem (format)

Packages: e2fsprogs (ext4), xfsprogs, btrfs-progs, exfatprogs, dosfstools. Install with apt, dnf, pacman -S or zypper install as usual.

4. Mount and add to /etc/fstab

# temporary mount
sudo mkdir -p /mnt/usb
sudo mount /dev/sdb1 /mnt/usb
# unmount
sudo umount /mnt/usb

If umount says the target is busy, a shell or program is using it; see umount: target is busy.

For a permanent mount, add a line to /etc/fstab. Format: device mountpoint type options dump pass.

# /etc/fstab examples
UUID=3f6c2a1e-9b0d-4e5f-8a7b-1c2d3e4f5a6b  /data       ext4   defaults,nofail                 0 2
UUID=5e4d3c2b-1a09-4f8e-9d7c-6b5a4f3e2d1c  /srv/xfs    xfs    defaults,nofail                 0 0
UUID=A1B2-C3D4                              /mnt/share  exfat  defaults,nofail,uid=1000,gid=1000 0 0
/swapfile                                   none        swap   sw                              0 0

• nofail lets the system boot if the disk is missing, instead of dropping to emergency mode. Use it for every non-essential disk.
• The last field (pass) is 1 for /, 2 for other ext4 filesystems to be checked at boot, 0 for XFS, Btrfs, FAT and network shares.
• FAT and exFAT have no Unix permissions, so uid=/gid= set who owns the files.

Always test before rebooting:

sudo findmnt --verify
sudo systemctl daemon-reload
sudo mount -a
findmnt /data

findmnt --verify catches typos and unknown UUIDs; mount -a mounts everything in fstab and errors out on bad lines. A bad fstab line is the most common cause of a system that boots to emergency mode.

5. Resize partitions and filesystems

Grow (most common: a VM disk was enlarged)

# 1. grow the partition to fill the disk (package: cloud-guest-utils / cloud-utils-growpart)
sudo growpart /dev/sda 2
# 2. grow the filesystem, online
sudo resize2fs /dev/sda2          # ext4
sudo xfs_growfs /                 # XFS: takes the mount point
sudo btrfs filesystem resize max / # Btrfs

Shrink ext4

Shrink the filesystem first, then the partition. It must be unmounted, so for / boot a live USB.

sudo umount /dev/sdb1
sudo e2fsck -f /dev/sdb1
sudo resize2fs /dev/sdb1 200G
# then shrink the partition to slightly more than 200G in parted/GParted, and run:
sudo resize2fs /dev/sdb1

GParted from a live USB does all of this in one step and is the safer choice for desktop users.

LVM (Ubuntu Server, Fedora Server, RHEL defaults)

# overview
sudo pvs; sudo vgs; sudo lvs
# use free space in the volume group and grow the filesystem in one step
sudo lvextend -r -l +100%FREE /dev/ubuntu-vg/ubuntu-lv
# add a new disk to the volume group first if it has no free space
sudo pvcreate /dev/sdc
sudo vgextend ubuntu-vg /dev/sdc

Ubuntu Server's installer often leaves half the volume group unallocated, which is why a new server shows a small /. The lvextend -r command above fixes that and is the usual answer to No space left on device on a fresh server.

6. Check disk health with SMART

sudo apt install smartmontools      # dnf install smartmontools / pacman -S smartmontools / zypper install smartmontools
# overall verdict
sudo smartctl -H /dev/sda
# full attributes
sudo smartctl -a /dev/sda
sudo smartctl -a /dev/nvme0
# start a short self-test (2 minutes), then read the result
sudo smartctl -t short /dev/sda
sudo smartctl -l selftest /dev/sda

What to look at:

The NVMe-specific tool gives the same data: sudo nvme smart-log /dev/nvme0 (package nvme-cli). Enable continuous monitoring with sudo systemctl enable --now smartd (the service may be called smartmontools on Debian/Ubuntu).

A filesystem that suddenly turns read-only is often the first symptom of a failing disk; see Read-only file system, and check sudo dmesg | grep -i -E 'error|i/o'.

7. Check and repair filesystems

# ext4: must be unmounted (use a live USB for /)
sudo umount /dev/sdb1
sudo fsck -f /dev/sdb1
# XFS
sudo xfs_repair /dev/sdb1
# Btrfs: scrub runs online and verifies checksums
sudo btrfs scrub start -B /
sudo btrfs device stats /

Never run fsck on a mounted filesystem; it can corrupt it. For the root filesystem on ext4, you can also schedule a check at next boot by adding fsck.mode=force to the kernel command line once (see kernel parameters).

Swap files

sudo fallocate -l 4G /swapfile
sudo chmod 600 /swapfile
sudo mkswap /swapfile
sudo swapon /swapfile
echo '/swapfile none swap sw 0 0' | sudo tee -a /etc/fstab
swapon --show

On Btrfs, create swap files with sudo btrfs filesystem mkswapfile --size 4G /swap/swapfile instead, because swap files there need copy-on-write disabled. Fedora uses zram (compressed RAM swap) by default, so a swap file is usually unnecessary there. The Arch Wiki's fstab and partitioning pages are good deeper references for any distro.

Frequently asked questions