Machine lost network after its first reboot

Stable reproduction on Ubuntu 20.04/22.04/24.04 LTS with MAAS 3.5.3 stable. I don’t know why, but it is a horrible experience to sit and look at the IPMI monitor and watch it for several power cycles.
I tried:

  • Release the machine.
  • Delete the machine from MAAS and re-add it to MAAS.
  • Do commissioning before deployment.
  • Set DHCP on eno1 or assign static IP address on eno1.

Similar to this post:


The only name server assigned to the machine is maas itself. So, this issue is not related to DNS settings.

No useful log on the maas side, because the machine failed to stream the log back to MAAS after reboot due to network disruption.

Wdym by “machine lost network config”? What’s inside /etc/netplan?

Hello r00ta,

“Machine lost network config”: after the first reboot during OS installation, the machine will lose network connection, it does not respond to any traffic on the allocated IP address. On the MAAS Web-UI, the machine is stuck at “Rebooting” and results in deployment failure due to timeout.

As for /etc/netplan, I don’t know as I cannot connect to the machine. User-customised cloud-init script is designed to be loaded after the reboot, so it does not help to set the user password in the cloud-init script since it couldn’t load due to network failure (it cannot connect to MAAS controller).

Do you have access to the machine console? Is there any error message there?

I have IPMI access and did a screen recording.
But unfortunately, I don’t know the default password set by MAAS, so I couldn’t sign in for debugging.

The video seems to be cut. Do you ever get to the console?

Yes – attached is the part 2 of the recording

I’d suggest to enable login with password for a user so to debug what went wrong on the machine itself.

Hi r00ta, how do I enable login with a password for a user?
The cloud-init script won’t work because it is expected to run after the first reboot during OS installation.

You can either create a custom ubuntu image with packer-maas where you enable a username with a password or you can unsquash the image under /var/snap/maas/common/maas/image-storage and you add an entry in the /etc/shadow (roughly, this is the hack to do it), then you create again the squashfs. You can try with some test environments first

1 Like