Alpine on Scaleway - Your IPv6 box for $0.15/Month

Moohr
November 21, 2025
3 min read

November 21, 2025

Please note that this is done for scientific purposes and for educating on how Linux servers works.

Delete the 10GB volume of storage. Go to local storage and add a 1GB volume.
Boot the VM in rescue mode.
Once inside, find your disk:

root@scw-strange-lewin:~# lsblk
NAME    MAJ:MIN RM   SIZE RO TYPE MOUNTPOINTS
sda       8:0    0   9.3G  1 disk
├─sda1    8:1    0   3.5G  1 part /media/root-ro
├─sda14   8:14   0     4M  1 part
├─sda15   8:15   0   106M  1 part /boot/efi
└─sda16 259:0    0   913M  1 part /media/root-ro/boot
vda     253:0    0 953.7M  0 disk
root@scw-strange-lewin:~#

In here, we can see that that vda is the 1GB disk we added. Let’s get Alpine onto that.

Download Alpine, DD, and set up some required mounts:

First let’s grab the cloud image

wget https://dl-cdn.alpinelinux.org/alpine/v3.22/releases/cloud/generic_alpine-3.22.2-x86_64-uefi-cloudinit-r0.qcow2

To actually DD it onto a blank block device (aka disk), we need to convert it from the qcow (qemu copy-on-write) format into raw. However we don’t actually have enough storage to store a second copy in the rescue environment, but we can just tell qemu-img to dump it directly to the disk.

apt update && apt install qemu-utils -y

#qemu-img convert -O raw generic_alpine-3.22.2-x86_64-uefi-cloudinit-r0.qcow2 alpine.raw
#dd if=alpine.raw of=/dev/vda bs=1M status=progress
#if you have enough storage in your rescue you can do it this way (old fashioned)
#otherwise you can just do it like this

qemu-img convert -O raw generic_alpine-3.22.2-x86_64-uefi-cloudinit-r0.qcow2 /dev/vda

sync

We have now written alpine to the disk image. It’ll boot, but we haven’t added any credentials, so it’s pretty useless. To do that, we need to perform some setup. Let’s expand the disk to full size first:

sgdisk -e /dev/vda
parted -s /dev/vda resizepart 2 100%
e2fsck -f /dev/vda2
resize2fs /dev/vda2

Afterwards, mount the disk and chroot into the system:

mkdir -p /mnt/alpine
mount /dev/vda2 /mnt/alpine

Then mount some required stuff from the rescue system to our new system:

mount -t proc none /mnt/alpine/proc
mount -t sysfs none /mnt/alpine/sys
mount --rbind /dev /mnt/alpine/dev
mount --make-rslave /mnt/alpine/dev

Enter alpine (hurrah)

chroot /mnt/alpine /bin/sh

Set up some stuff.

Let’s add the ssh key in.

mkdir -p /root/.ssh && echo "ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIFWnkfCZETwETd8bpdiz2/RRvsswdc+ryNZK5jXGi7jf" > /root/.ssh/authorized_keys

Then we need to add these two lines to the start of /etc/ssh/sshd_config

PermitRootLogin prohibit-password
PubkeyAuthentication yes

Then run this to remove the locked root account

passwd -d root

Then you’re basically done. If you find the boot times annoyingly long, you can disable the cloudinit:

touch /etc/cloud/cloud-init.disabled

Do keep in mind you will have to set up your own IPs. This shouldn’t be a huge problem, because you have VNC access to the machine in Scaleway’s panel.

cat <<EOF > /etc/network/interfaces
auto lo
iface lo inet loopback

auto eth0
iface eth0
    address 2001:bc8:1640:5a75:dc00:ff:fe60:1a21/64
    gateway fe80::dc00:ff:fe60:1a22

auto dummy0
iface dummy0
    link-type dummy
    address 172.23.77.209/32
    address fd09:ec7b:bc88:209::1/128

auto dummy2
iface dummy2
    link-type dummy
    address 172.23.77.199/32
    address fd09:ec7b:bc88:199::1/128
EOF

cat <<EOF > /etc/resolv.conf
nameserver 2606:4700:4700::1111
nameserver 2606:4700:4700::1001
EOF

Tags:
Vps