Skip to content

Instantly share code, notes, and snippets.

@craftyc0der

craftyc0der/README.md

Last active April 22, 2024 18:31
Show Gist options
  • Star 11 You must be signed in to star a gist
  • Fork 2 You must be signed in to fork a gist
  • Save craftyc0der/3cbdb1f9ed60aa94f8cfc0f54b719b95 to your computer and use it in GitHub Desktop.
Save craftyc0der/3cbdb1f9ed60aa94f8cfc0f54b719b95 to your computer and use it in GitHub Desktop.
Download ZIP
Setup Snapper on Fedora 34
Raw
README.md

How to setup RAID1 with BTRFS on Fedora

Here we list our disks for future use.

sudo lsblk
NAME                                          MAJ:MIN RM   SIZE RO TYPE  MOUNTPOINT
zram0                                         252:0    0     8G  0 disk  [SWAP]
nvme1n1                                       259:0    0 953.9G  0 disk
├─nvme1n1p1                                   259:1    0   190M  0 part  /boot/efi
├─nvme1n1p3                                   259:3    0     1G  0 part  /boot
└─nvme1n1p7                                   259:7    0 932.6G  0 part
  └─luks-13f88890-0b52-4080-a3b6-b406d616c659 253:0    0 932.6G  0 crypt /home
nvme2n1                                       259:8    0 931.5G  0 disk
nvme0n1                                       259:10   0 476.9G  0 disk
└─nvme0n1p1                                   259:11   0 476.9G  0 part

Create second disk partitions using parted & Disks

Duplicate the /boot and /boot/efi partition structure of disk one. sudo parted /dev/nvme2n1 Use Disks to create the btrfs with LUKS

Here we list our disks for future use.

sudo lsblk
NAME                                          MAJ:MIN RM   SIZE RO TYPE  MOUNTPOINT
zram0                                         252:0    0     8G  0 disk  [SWAP]
nvme1n1                                       259:0    0 953.9G  0 disk
├─nvme1n1p1                                   259:1    0   190M  0 part  /boot/efi
├─nvme1n1p2                                   259:3    0     1G  0 part  /boot
└─nvme1n1p3                                   259:7    0 932.6G  0 part
  └─luks-13f88890-0b52-4080-a3b6-b406d616c659 253:0    0 932.6G  0 crypt /var/lib/docker/btrfs
nvme2n1                                       259:8    0 931.5G  0 disk
├─nvme2n1p1                                   259:1    0   190M  0 part
├─nvme2n1p2                                   259:3    0     1G  0 part
└─nvme2n1p3                                   259:9    0 931.5G  0 part
  └─luks-429c570c-2743-4cc3-beaa-dfc8facb118c 253:1    0 931.5G  0 crypt
nvme0n1                                       259:10   0 476.9G  0 disk
└─nvme0n1p1                                   259:11   0 476.9G  0 part

Add second disk to path

sudo btrfs device usage /
/dev/mapper/luks-13f88890-0b52-4080-a3b6-b406d616c659, ID: 1
   Device size:           932.57GiB
   Device slack:              0.00B
   Data,single:            85.01GiB
   Metadata,single:         8.01GiB
   System,single:           4.00MiB
   Unallocated:           839.55GiB

We will have to add the -f option to force replacement of the filesystem on the disk if it has an existing filesystem. This is destructive!

I also takes a long time if there is a lot of data. Be patient. You can follow progress by running the usage command from a different terminal.

sudo btrfs device add -f /dev/mapper/luks-429c570c-2743-4cc3-beaa-dfc8facb118c /
sudo btrfs device usage /
sudo btrfs balance start -dconvert=raid1 -mconvert=raid1 /
sudo btrfs device usage /

Update crypttab

Add to /etc/crypttab:

luks-429c570c-2743-4cc3-beaa-dfc8facb118c UUID=429c570c-2743-4cc3-beaa-dfc8facb118c none discard

Update Bootloader

Now we need to add the luks to grub.

Add rd.luks.uuid=luks-429c570c-2743-4cc3-beaa-dfc8facb118c to the GRUB_CMDLINE_LINUX value in /etc/default/grub

Then rebuild grub.

Then rebuild initramfs.

sudo -i
nano /etc/default/grub
grub2-mkconfig -o "$(readlink /etc/grub2.cfg)"
dracut --force
reboot

Duplicate /boot to the second disk as a precaution

sudo dd if=/dev/nvme1n1p2 of=/dev/nvme2n1p2 bs=1024 status=progress
sudo dd if=/dev/nvme1n1p1 of=/dev/nvme2n1p1 bs=1024 status=progress

How to user Snapper on Fedora 34

Snapper is a snapshot management tool for btrfs. It is available here. http://snapper.io/

Here we list our top level subvolumes.

If you are using Docker's btrfs volume driver you end up with a lot snapshots which necessitates the grep filter used below.

sudo btrfs subvolume list / | grep "level 5"
ID 256 gen 1987943 top level 5 path home
ID 257 gen 1987942 top level 5 path root

Install snapper

This installs snapper and an automatic snapshot before and after each dnf install.

sudo dnf install snapper python-dnf-plugin-snapper

Setup snapper

Here we create a snapper config for the root subvolume.

sudo snapper -c root create-config /

It is not stored at the top level however, so we will move the subvolume from /.snapshots (within the root subvolume) to level 5. We want it at the top level so we can rollback our root partition and boot into it directly.

sudo btrfs subvolume list / | grep snapshots
sudo btrfs subvolume delete /.snapshots

Now we will make a new folder within the root path and create a top level subvolume and mount it there.

sudo mkdir /.snapshots
sudo mkdir /backups
sudo mkdir /mnt/btrfs
sudo mount /dev/dm-0 -o subvolid=5 /mnt/btrfs
cd /mnt/btrfs
sudo btrfs subvolume create snapshots
sudo btrfs subvolume create backups
cd ..
sudo umount /mnt/btrfs
sudo rmdir btrfs/

There are now two new top level subvolume.

sudo btrfs subvolume list / | grep "level 5"
ID 256 gen 1987943 top level 5 path home
ID 257 gen 1987942 top level 5 path root
ID 10963 gen 1987638 top level 5 path snapshots
ID 10963 gen 1987638 top level 5 path backups

Mount the new subvolumes

sudo nano /etc/fstab

We add a new lines at the bottom which mounts the new subvolumes.

UUID=2402f445-7fd2-4a8b-8b53-6c27c67fb58d /                       btrfs   subvol=root,compress=zstd:1,x-systemd.device-timeout=0 0 0
UUID=13f88890-0b52-4080-a3b6-b406d616c659 /boot                   ext4    defaults        1 2
UUID=62A1-49D9          /boot/efi               vfat    umask=0077,shortname=winnt 0 2
UUID=2402f445-7fd2-4a8b-8b53-6c27c67fb58d /home                   btrfs   subvol=home,compress=zstd:1,x-systemd.device-timeout=0 0 0
UUID=2402f445-7fd2-4a8b-8b53-6c27c67fb58d /.snapshots             btrfs	  subvol=snapshots,x-systemd.device-timeout=0 0 0
UUID=2402f445-7fd2-4a8b-8b53-6c27c67fb58d /backups                btrfs	  subvol=backups,x-systemd.device-timeout=0 0 0

Let's mount it.

sudo mount -a

We will make user folder with user permissions.

sudo mkdir /backups/$USER
sudo chown $USER:$USER /backups/$USER
cd $HOME
ln -s /backups/$HOME backups

Integrate Snapper with Grub to allow / rollbacks

The default subvolume is just top level. This is not specific enough for our needs. We want to be able to set this explicitly to a subvolume ID.

sudo btrfs subvolume get-default /
ID 5 (FS_TREE)

Recall above our root subvolume ID was 257 we will set that to our default.

sudo btrfs subvolume set-default 257 /
sudo btrfs subvolume get-default /
ID 257 gen 1988138 top level 5 path root

Now we need to modify the Grub config.

sudo grubby --info=ALL
...
index=2
kernel="/boot/vmlinuz-5.13.14-200.fc34.x86_64"
args="ro rootflags=subvol=root rhgb quiet"
...

We want to remove the rootflags=subvol=root argument so it will honor rollbacks requested by snapper.

sudo grubby --update-kernel=ALL --remove-args="rootflags=subvol=root"
sudo grubby --info=ALL
...
index=2
kernel="/boot/vmlinuz-5.13.14-200.fc34.x86_64"
args="ro rhgb quiet"
...

All better, so now we reboot.

reboot

Rollback to a root snapshot

sudo snapper ls
 # | Type   | Pre # | Date            | User | Cleanup |                                        | Userdata
---+--------+-------+-----------------+------+---------+----------------------------------------+---------
0  | single |       |                 | root |         | current                                |
1  | pre    |       | Sun 18 Sep 2021 | root | number  | /bin/dnf -y install ...fc34.x86_64.rpm |
2  | post   |     1 | Sun 19 Sep 2021 | root | number  | /bin/dnf -y install ...fc34.x86_64.rpm |

Oh no, our install broke everything!

sudo snapper --ambit classic rollback 1
reboot

Life is good!

Add automatic snapshots of home subvolume

We will create a snapper config for /home and give our personal user access since this is a single user PC.

sudo snapper -c home create-config /home
sudo snapper -c home set-config SYNC_ACL=yes ALLOW_USERS=$USER
sudo snapper list-configs
Config | Subvolume
-------+----------
home   | /home
root   | /

Make our first home snapshot

Notice sudo is not required anymore.

snapper -c home create --description "First Snapshot"
snapper -c home ls
 # | Type   | Pre # | Date            | User | Cleanup  | Description    | Userdata
---+--------+-------+-----------------+------+----------+----------------+---------
0  | single |       |                 | root |          | current        |
1  | single |       | Sat 18 Sep 2021 | josh |          | First Snapshot |

Setup automatic backups

sudo nano /etc/snapper/configs/root

Disable automatic hourly snapshots of root.

# create hourly snapshots
TIMELINE_CREATE="no"

# cleanup hourly snapshots after some time
TIMELINE_CLEANUP="not"
sudo nano /etc/snapper/configs/home

Configure automatic hourly snapshots of home.

# create hourly snapshots
TIMELINE_CREATE="yes"

# cleanup hourly snapshots after some time
TIMELINE_CLEANUP="yes"

# limits for timeline cleanup
TIMELINE_MIN_AGE="1800"
TIMELINE_LIMIT_HOURLY="5"
TIMELINE_LIMIT_DAILY="7"
TIMELINE_LIMIT_WEEKLY="0"
TIMELINE_LIMIT_MONTHLY="1"
TIMELINE_LIMIT_YEARLY="0"

Turn on timer for automatic snapshots.

sudo systemctl enable --now snapper-timeline.timer
sudo systemctl enable --now snapper-cleanup.timer

After waiting an hour...

snapper -c home ls
 # | Type   | Pre # | Date            | User | Cleanup  | Description    | Userdata
---+--------+-------+-----------------+------+----------+----------------+---------
0  | single |       |                 | root |          | current        |
1  | single |       | Sat 18 Sep 2021 | josh |          | First Snapshot |
2  | single |       | Sun 18 Sep 2021 | root | timeline | timeline       |
@paraduxos
Copy link

Thank you very much :)

@JosiahBull
Copy link

cheers :)

@constantin-diaconu
Copy link

I get the following message when using sudo mount /dev/dm-0 -o subvolid=5 /mnt/btrfs:

mount: /mnt/btrfs: special device /dev/dm-0 does not exist.

I am fairly new to fedora and Linux in general and I'm not sure what it means or what to do about it.

@xenithorb
Copy link

xenithorb commented May 6, 2022 •
edited

Thank you for the excellent walkthrough. I would appreciate clarification on a few items:

  1. Ambit classic, why was this a chosen option?
    The man page states:

    In other ambits than classic the rollback command does what is required to do a rollback. Anyway it is recommended to use specific programs in that case.

  2. What is the backups directory used for? It doesn't seem to be a function of snapper that I can tell.

@CristianArean
Copy link

Thanks for the tutorial, I'm new at snapper and I found it really helpful. If I have not misunderstood the tutorial, this allows me to restore snapshot by using the grub. Would you like to show how that is done?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment