Cloud-init is the industry standard multi-distribution method for cross-platform cloud instance initialization. It is supported across all major public cloud providers, provisioning systems for private cloud infrastructure, and bare-metal installations.

Using Proxmox templates in tandem with Cloud-init streamlines the process of launching new VMs. With this approach, a template can rapidly generate a new VM, and Cloud-init takes care of the initial setup during boot, reducing your tasks to simply setting up the hostname and initial user account. This eliminates the need for manually configuring a fresh operating system installation for each VM. This guide focuses on creating a template Debian configuration.

Although Debian doesn’t offer a dedicated image specifically for this scenario, the Debian images intended for OpenStack/Cloud are equipped with Cloud-init compatibility. For more information on how Proxmox supports Cloud-Init, refer to Proxmox’s documentation.

Assuming you’re a beginner, running through this guide to get a template set up and configured should take you approximately 10 minutes. Diving a bit deeper in to each of the commands will take longer to understand what’s happening behind the scenes. It can be useful to have the Proxmox web interface open as you proceed through each step so you can see how the changes after each command are reflected in the web interfac.

Good luck and happy hacking!

Download the Cloud-init image

Download the latest genericcloud image from the Debian Official Cloud Images Repository directly on to the Proxmox host

Debian 11

1

wget https://cloud.debian.org/images/cloud/bullseye/20231004-1523/debian-11-genericcloud-amd64-20231004-1523.qcow2

Debian 12

1

wget https://cloud.debian.org/images/cloud/bookworm/latest/debian-12-genericcloud-amd64.qcow2

Create a new VM using the image

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25

# Set the VM ID to operate on
VMID=9001
# Choose a name for the VM
TEMPLATE_NAME=Debian12CloudInit
# Choose the disk image to import
DISKIMAGE=debian-12-genericcloud-amd64.qcow2
# Select Host disk
HOST_DISK=local-zfs

# Create the VM
qm create $VMID --name $TEMPLATE_NAME --net0 virtio,bridge=vmbr0
# Set the OSType to Linux Kernel 6.x
qm set $VMID --ostype l26
# Import the disk
qm importdisk $VMID $DISKIMAGE $HOST_DISK
# Attach disk to scsi bus
qm set $VMID --scsihw virtio-scsi-pci --scsi0 $HOST_DISK:vm-$VMID-disk-0
# Set scsi disk as boot device
qm set $VMID --boot c --bootdisk scsi0
# Create and attach cloudinit drive
qm set $VMID --ide2 $HOST_DISK:cloudinit
# Set serial console, which is needed by OpenStack/Proxmox
qm set $VMID --serial0 socket --vga serial0
# Enable Qemu Guest Agent
qm set $VMID --agent enabled=1 # optional but recommened

Configure other VM options to suite needs

1
2

# Start the VM at boot
qm set $VMID --onboot 1

Convert the VM into a template

1

qm template $VMID

Create a new VM from the template

1
2
3
4

# Set a variable for our new VM ID
NEW_VMID=2001
NEW_VM_NAME=MyNewServer
qm clone $VMID $NEW_VMID --name $NEW_VM_NAME --full

Configure the VM

Once the VM is created, there are some common adjustments we might want to make, such as assigning additional vCPUs, RAM, or disk space.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

# Show Current configs
qm config $NEW_VMID

# Allocate 2GB of RAM
qm set $NEW_VMID --memory 2048
# Allocate 2 vCPU Cores
qm set $NEW_VMID --cores 2
# Resize bootdisk
## Get bootdisk interface
NEW_VM_BOOTDISK_INTERFACE=$(qm config $NEW_VMID | grep bootdisk | awk '{print $2}')
## Resize bootdisk to 32GB
qm disk resize $NEW_VMID $NEW_VM_BOOTDISK_INTERFACE 32G

Configure Cloud-init

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11

# Set Network config
## Static
qm set $NEW_VMID --ipconfig0 ip=172.17.1.132/24,gw=172.17.1.1
## DHCP
qm set $NEW_VMID --ipconfig0 ip=dhcp

# Set Username
qm set $NEW_VMID --ciuser aiden

# Setup public SSH keys (one key per line, OpenSSH format).
qm set $NEW_VMID --sshkeys ~/vm_sshkeys/keylist

Boot and initialise the VM

1
2
3
4
5
6
7

sudo apt update
sudo apt upgrade

# Install and enable QEMU guest agent
sudo apt install qemu-guest-agent
sudo systemctl enable qemu-guest-agent
sudo systemctl start qemu-guest-agent

References