VM doesn't shut down when trying to destroy about terraform-provider-proxmox HOT 10 CLOSED

Hi @MichalMarchewka 👋🏼

From the example tf template, it doesn't looks like your VM has any OS to load. I think it is just looping though the BIOS boot sequence trying to bootstrap, but there is no bootloader available. You can check that on the VM console.

The provider can't do anything in this case, neither Proxmox UI. You can directly issue a qemu command quit via the Monitor section in the UI to terminate such VMs:

from terraform-provider-proxmox.

Hello @bpg 👋🏻
Thank you for the quick reply. To confirm - I can shut down the VM using the PVE interface and then delete it manually. I'm just wondering if it's possible in your provider to force the virtual machine to shut down after removing it from the TF code and then delete it. I'm very new to Proxmox, I've only been playing with it for two days, but this kind of thing works in the ESXi and vSphere TF provider (even if no image isattached to the VM instance).

from terraform-provider-proxmox.

I've added following code to VM configuration:

cdrom {
    enabled = true
    file_id = "local:iso/ubuntu_server_au.iso"
}

VM started and ISO was loaded. When I remove the VM config from TF it's the same behaviour as before, but system is installing (no looped boot sequence this time):

1. PVE UI says the VM is shuting down
2. TF console says it keeps destroying

from terraform-provider-proxmox.

Interesting... 🤔 I just want to confirm, that Shutdown from the UI works for you, but when provider does that, it hangs?

from terraform-provider-proxmox.

Well, not exectly to be honest. I'm not able to shutdown VM from the UI but I'm able to stop it. After I stop it manually then remove from TF, terraform apply --auto-approve successfully removes it from Proxmox.

from terraform-provider-proxmox.

resource "proxmox_virtual_environment_vm" "test_01" {
  name        = "test-01"
  description = "Managed by Terraform"
  tags        = ["test"]
  on_boot     = true
  node_name = "pve"

  #agent {
  #  enabled = true
  #}

  operating_system {
    type = "l26"
  }

  cpu {
    cores = 2
    type  = "host"
    numa  = true
  }

  memory {
    dedicated = 1024
  }

  disk {
    size         = "30"
    interface    = "virtio0"
    datastore_id = "local-lvm"
    file_format  = "raw"
  }

  network_device {
    bridge = "vmbr0"
    model  = "virtio"
  }

  cdrom {
    enabled = true
    file_id = "local:iso/ubuntu_server_au.iso"
  }

  initialization {
    # interface = "scsi1"
    ip_config {
      ipv4 {
        address = "dhcp"
      }
    }
  }
}

Scenario 1:

1. Create VM resource
2. Let it start
3. Remove the resource from TF and apply
4. Proxmox UI says it's shutting down, TF says it's still destroying

Scenario 2:

1. Create VM resource
2. Let it start
3. Stop it manually via Proxmox UI
4. Remove the resource from TF and apply
5. VM is removed from Proxmox

from terraform-provider-proxmox.

Ah... then stop_on_destroy should help 🤞🏼

from terraform-provider-proxmox.

Amazing! It's my fault I missed this argument. Thank you!

from terraform-provider-proxmox.

No prob! 🙂 Please let me know if it works for you.

from terraform-provider-proxmox.

I assume it worked :)

from terraform-provider-proxmox.