Skip to content

Proxmox Dev

🚧 UNDER CONSTRUCTION 🚧

this is an experimental approach, i am still building this workflow 😀

📸 Summary

This workflow deploys the dev environment for proxmox, this allows testing of tofu automation before deploying to prod.

🗃️ Repo

git clone git@gitlab.com:loganmancuso_public/infrastructure/local/proxmox-dev.git

📜 Dependencies

📃 Sample TFvars

instance_settings = {
  name      = "proxmox-dev"
  ip        = "XXX.XXX.XXX.XXX"
  disk_size = 60
  iso       = "proxmox-ve_8.1-2.iso"
}

⚙️ Deployment Instructions

🛑 Pre-Deployment

configure your local user to deploy vms by adding them to the libvirt group. https://lunux.net/use-virt-manager-as-a-non-root-user-on-linux/ https://github.com/dmacvicar/terraform-provider-libvirt/issues/978#issuecomment-1276244924

sudo usermod -a -G libvirt $(whoami)
newgrp libvirt
# to uncomment these lines in the conf 
sudo sed -i 's/^# *\(unix_sock_group = "libvirt"\)/\1/' /etc/libvirt/libvirtd.conf
sudo sed -i 's/^# *\(unix_sock_rw_perms = "0770"\)/\1/' /etc/libvirt/libvirtd.conf
sudo sed -i 's/#security_driver = "selinux"/security_driver = "none"/' /etc/libvirt/qemu.conf

sudo chown -R libvirt-qemu:libvirt /var/lib/libvirt/images/
# sudo chmod 770 /var/lib/libvirt/images
sudo chown -R libvirt-qemu:libvirt /var/lib/virt/images/

download the proxmox iso and put it into a folder at the root called iso I was going to add in the download of the iso to the workflow but its too variable. I would also need to put a life-cycle prevent destroy on the iso resource and would make the entire workflow more difficult to use as I am constantly spinning up and tearing down this vm for testing. You will also need to copy that iso to /var/lib/libvirt/images/ in order for libvirt to have permissions to it. 

rsync -v iso/*.iso /var/lib/libvirt/images/

🟢 Deployment

to deploy this workflow link the environment tfvars folder to the root directory.

ln -s env/* .
tofu init .
tofu plan
tofu apply

🏁 Post-Deployment

after this proxmox node is stood up you will need to do a few more things

  1. copy ssh key to the proxmox node ssh-copy-id -i ~/.ssh/id_ed25519 root@100.64.0.10
  2. change the root password in the gui
  3. you will need to run the Datacenter deployment on this newly deploy proxmox node and follow the post deployment instructions there as well.

📝 Notes

📅 Tasks

👎 Known Issues

  • the instance ip will not be available on first deploy with dhcp but a terraform refresh will fix the issue.
  • issue deleting the storage pool on a bad deployment with cloud_init. when deploying if vm becomes in an unusable state when deploying try these commands to clear the state and start fresh. 
    sudo virsh pool-list
sudo virsh pool-destroy default
sudo virsh pool-destroy Downloads
sudo virsh pool-undefine proxmox-dev
sudo virsh undefine proxmox-dev
sudo virsh domstate proxmox-dev
  • to destroy just the instance use this command tf destroy --target libvirt_domain.instance

expand disk for the vm

qemu-img resize --preallocation=full /var/lib/virt/images/proxmox-dev/os_image_proxmox +XG

<<Local Infrastructure --- Datacenter>>