mkfs (Creating Filesystems)

# -----------------------------------------------
#  Creating a filesystem (mkfs)
# -----------------------------------------------

# mkfs.ext4 {device}
#   → Creates an ext4 filesystem
#   → A standard, general-purpose filesystem
#   Example: sudo mkfs.ext4 /dev/sdb1

# mkfs.ext4 -L {label} {device}
#   → Creates an ext4 filesystem with a volume label
#   Example: sudo mkfs.ext4 -L data01 /dev/sdb1

# mkfs.xfs {device}
#   → Creates an xfs filesystem
#   → Optimized for large files and high I/O; the default on RHEL-based systems
#   Example: sudo mkfs.xfs /dev/sdb1

# mkfs.xfs -L {label} {device}
#   → Creates an xfs filesystem with a volume label
#   Example: sudo mkfs.xfs -L data01 /dev/sdb1

# mkfs.btrfs {device}
#   → Creates a btrfs filesystem
#   → Supports snapshots, subvolumes, and transparent compression
#   Example: sudo mkfs.btrfs /dev/sdb1

# mkfs.btrfs -L {label} {device}
#   → Creates a btrfs filesystem with a volume label
#   Example: sudo mkfs.btrfs -L data01 /dev/sdb1

# -----------------------------------------------
#  Mounting and unmounting (mount / umount)
# -----------------------------------------------

# mount {device} {mount point}
#   → Mounts a filesystem at a directory
#   Example: sudo mount /dev/sdb1 /mnt/data

# mount -t {filesystem type} {device} {mount point}
#   → Mounts a filesystem with the type specified explicitly
#   Example: sudo mount -t ext4 /dev/sdb1 /mnt/data
#   Example: sudo mount -t xfs /dev/sdb1 /mnt/data

# mount -o {options} {device} {mount point}
#   → Mounts a filesystem with the given mount options
#   Example: sudo mount -o ro /dev/sdb1 /mnt/data        (read-only)
#   Example: sudo mount -o remount,rw /dev/sdb1 /mnt/data (remount as read-write)

# mount
#   → Lists all currently mounted filesystems

# umount {mount point or device}
#   → Unmounts a filesystem
#   Example: sudo umount /mnt/data
#   Example: sudo umount /dev/sdb1

# -----------------------------------------------
#  /etc/fstab format
# -----------------------------------------------

# {device}  {mount point}  {FS type}  {options}  {dump}  {pass}
#
#  Field 1 (device) : Device file, UUID, or LABEL
#  Field 2 (mount point) : Directory path where the filesystem is mounted
#  Field 3 (FS type) : ext4 / xfs / btrfs / swap, etc.
#  Field 4 (options) : defaults / ro / noatime / nofail, etc.
#  Field 5 (dump) : Whether to include in dump backups (usually 0)
#  Field 6 (pass) : fsck check order (root = 1, others = 2 or 0)
#
#  Example (UUID-based):
#  UUID=a1b2c3d4-e5f6-7890-abcd-ef1234567890 /mnt/data  ext4  defaults,nofail  0  2

Command / Field	Description
mkfs.ext4 {device}	Creates an ext4 filesystem. It is the standard choice for general-purpose use and the default on Ubuntu and Debian. Journaling makes recovery after a crash straightforward.
mkfs.xfs {device}	Creates an xfs filesystem. It performs well with large files and high-concurrency I/O, and is the default on RHEL-based Linux distributions. An existing xfs filesystem can be expanded online, but cannot be shrunk.
mkfs.btrfs {device}	Creates a btrfs filesystem. It is a next-generation filesystem with built-in snapshots, subvolumes, transparent compression, and RAID support. It is the default on Fedora and openSUSE.
mount -t {type} {device} {mount point}	Mounts a device with the filesystem type specified explicitly. The type can be omitted for auto-detection, but specifying it avoids misidentification.
umount {mount point}	Unmounts a mounted filesystem. If a process is actively using the target directory, you will get a "device is busy" error. Use lsof +D {mount point} to find the processes accessing it.
mount -a	Mounts all entries listed in /etc/fstab. Use this to verify that a fstab edit is correct before rebooting.
blkid {device}	Displays the UUID, label, and filesystem type of a device. Use this to get the UUID when writing a fstab entry.
fstab field 1 (device)	Specifies the device to mount. Device paths such as /dev/sdb1 can change after a reboot, so using UUID=... or LABEL=... is recommended.
fstab field 2 (mount point)	Specifies the directory path where the filesystem will be mounted. The directory must be created with mkdir before mounting.
fstab field 3 (FS type)	Specifies the filesystem type. Accepted values include ext4, xfs, btrfs, swap, and vfat. Use auto to let the system detect the type automatically.
fstab field 4 (options)	Specifies mount options as a comma-separated list. defaults (rw / suid / exec / auto / nouser / async) is the typical starting point. nofail allows the system to continue booting even if the device is not present.
fstab field 5 (dump)	Indicates whether the filesystem should be included in backups by the legacy dump command. On modern systems this is normally set to 0 (disabled).
fstab field 6 (pass)	Specifies the order in which fsck checks the filesystem at boot. Set to 1 for the root filesystem, 2 for additional disks, and 0 to skip the check.

Creating and mounting an ext4 filesystem

# -----------------------------------------------
#  Format a new disk /dev/sdb with ext4 and mount it persistently
#  (Scenario: a disk for storing surveillance data for PSB Division 1)
# -----------------------------------------------

# Check the current state of the device
# Confirm that /dev/sdb has no partitions yet
sudo fdisk -l /dev/sdb

# Create a partition (use fdisk to create partition /dev/sdb1)
# See the disk_partition page for details
sudo fdisk /dev/sdb

# Create an ext4 filesystem
# Use -L to assign the volume label "psychopass-data"
sudo mkfs.ext4 -L psychopass-data /dev/sdb1

# Create the mount point directory
sudo mkdir -p /mnt/psychopass-data

# Temporarily mount the filesystem to verify it works
sudo mount /dev/sdb1 /mnt/psychopass-data

# Confirm the mount succeeded
df -h /mnt/psychopass-data

# Get the UUID to use in /etc/fstab for persistent mounting
# UUID uniquely identifies the device across reboots, unlike /dev/sdb1
sudo blkid /dev/sdb1

Run the following command:

$ sudo blkid /dev/sdb1
/dev/sdb1: LABEL="psychopass-data" UUID="c7a94e12-3f8b-4d02-9e5a-1b8d06f72c34" BLOCK_SIZE="4096" TYPE="ext4" PARTUUID="8e3d21a0-01"

The following example demonstrates this:

# Append the mount entry to /etc/fstab
# Use UUID to identify the device (device names can change)
# nofail prevents the boot process from halting if the disk is missing
echo 'UUID=c7a94e12-3f8b-4d02-9e5a-1b8d06f72c34 /mnt/psychopass-data  ext4  defaults,nofail  0  2' | sudo tee -a /etc/fstab

# Unmount and remount via fstab
sudo umount /mnt/psychopass-data

# Mount all fstab entries (any errors in fstab will surface here)
sudo mount -a

# Final check to confirm the filesystem is mounted
df -h /mnt/psychopass-data

Run the following command:

$ df -h /mnt/psychopass-data
Filesystem      Size  Used Avail Use% Mounted on
/dev/sdb1       197G  1.6G  186G   1% /mnt/psychopass-data

Creating an xfs filesystem (RHEL-based systems)

# -----------------------------------------------
#  Format /dev/sdc1 with xfs and mount it persistently
#  (Scenario: a high-I/O disk for MWPSB crime coefficient analysis logs)
# -----------------------------------------------

# Create an xfs filesystem
# xfs is the default format on RHEL-based systems (AlmaLinux / Rocky Linux)
sudo mkfs.xfs -L dominator-log /dev/sdc1

# Create the mount point
sudo mkdir -p /mnt/dominator-log

# Mount the filesystem, specifying xfs as the type explicitly
sudo mount -t xfs /dev/sdc1 /mnt/dominator-log

# Get the UUID and append it to fstab
sudo blkid /dev/sdc1

Run the following command:

$ sudo blkid /dev/sdc1
/dev/sdc1: LABEL="dominator-log" UUID="f2e85b31-7c4a-48d9-b16e-9a2f0c83e471" BLOCK_SIZE="512" TYPE="xfs" PARTUUID="1d7c9b44-01"

The following example demonstrates this:

# Append to fstab (xfs supports online expansion but cannot be shrunk)
echo 'UUID=f2e85b31-7c4a-48d9-b16e-9a2f0c83e471 /mnt/dominator-log  xfs  defaults,nofail  0  2' | sudo tee -a /etc/fstab

# Remount via fstab and verify
sudo umount /mnt/dominator-log
sudo mount -a
mount | grep dominator-log

Run the following command:

$ mount | grep dominator-log
/dev/sdc1 on /mnt/dominator-log type xfs (rw,relatime,attr2,inode64,logbufs=8,logbsize=32k,noquota)

Creating a btrfs filesystem (with snapshot support)

# -----------------------------------------------
#  Format /dev/sdd1 with btrfs and mount it with subvolumes
#  (Scenario: case file storage and snapshot management for the Public Safety Bureau)
# -----------------------------------------------

# Create a btrfs filesystem
sudo mkfs.btrfs -L sibyl-archive /dev/sdd1

# Create the mount point and mount the filesystem
sudo mkdir -p /mnt/sibyl-archive
sudo mount /dev/sdd1 /mnt/sibyl-archive

# Create btrfs subvolumes
# Snapshots can be managed independently per subvolume
sudo btrfs subvolume create /mnt/sibyl-archive/@data
sudo btrfs subvolume create /mnt/sibyl-archive/@snapshots

# List the btrfs subvolumes
sudo btrfs subvolume list /mnt/sibyl-archive

Run the following command:

$ sudo btrfs subvolume list /mnt/sibyl-archive
ID 256 gen 7 top level 5 path @data
ID 257 gen 8 top level 5 path @snapshots

The following example demonstrates this:

# Get the UUID and append to fstab
# Use the subvol= option to specify which subvolume to mount
sudo blkid /dev/sdd1

Run the following command:

$ sudo blkid /dev/sdd1
/dev/sdd1: LABEL="sibyl-archive" UUID="8a3d6f09-c21e-4b77-a9e3-5c0871d4b293" UUID_SUB="d4e5f6a7-b8c9-4d0e-1f2a-3b4c5d6e7f80" BLOCK_SIZE="4096" TYPE="btrfs" PARTUUID="2a4b6c8e-01"

The following example demonstrates this:

# Unmount and remount using the fstab entry with the subvolume specified
sudo umount /mnt/sibyl-archive

# Mount the @data subvolume at /mnt/sibyl-archive
echo 'UUID=8a3d6f09-c21e-4b77-a9e3-5c0871d4b293 /mnt/sibyl-archive  btrfs  defaults,subvol=@data,compress=zstd,nofail  0  0' | sudo tee -a /etc/fstab
sudo mount -a

# Show detailed information about the btrfs filesystem
sudo btrfs filesystem show /mnt/sibyl-archive

Run the following command:

$ sudo btrfs filesystem show /mnt/sibyl-archive
Label: 'sibyl-archive'  uuid: 8a3d6f09-c21e-4b77-a9e3-5c0871d4b293
	Total devices 1 FS bytes used 1.02MiB
	devid    1 size 500.00GiB used 3.02GiB path /dev/sdd1

The mkfs commands and mount are the fundamental operations for putting storage to use on Linux. Choosing the right filesystem type for the job is important. ext4 has a long track record and is the default on Debian and Ubuntu, making it a solid choice for general-purpose servers. xfs excels at large-file and high-concurrency I/O workloads and is the default on RHEL-based systems, making it well-suited for database and storage servers. btrfs offers advanced features such as snapshots, subvolumes, and transparent compression, enabling backup strategies to be built directly into the storage layer. Before creating a filesystem, always prepare a partition first using partition management (fdisk / parted). For flexible management of large disks, combining filesystems with LVM (Logical Volume Management) is also effective. After mounting, use the df / du commands to check disk usage. An incorrect /etc/fstab entry can prevent the system from booting, so always run mount -a to verify the configuration before rebooting.