Im trying to follow the http://www.ubuntu.com/download/cloud/install-openstack-with-autopilot to try out Autopilot and OpenStack on my own hardware.

I managed to get the MAAS Server up and running. It has two nics, one connected to LAN with internet connectivity( network), and the other NIC connected to a private switch( network).

I configured DHCP and DNS for the NIC2 on the MAAS Server (IP, set the ranges and added as Route for the DHCP Clients.

When I try to add my first Node, the PEX Boot works fine, It gets an IP from the DHCP server ( and starts loading the image I imported via the Images page for the MAAS Server(Ubuntu 14.04.1).

The new Node never shows up in the Nodes view on the MAAS Server, the Node installs, but does not shutdown(as I understand its suppose to do after success), and never shows up in the list of Nodes. It seems that the Node cant find it's Meta-Data. On the new node in terminal it shows.

cc_final_message.py[WARNING]: Used fallback datasource

During install, it also trows a 404 when trying to get its meta-data trough the url

url_helper.py[WARNING]: Calling '' failed : bas status code [404]

I have searched high and low for any information about this, can't find anything that seems relevant.

This "Lunch break install" is not looking much like it :)

Any help much appreciated.


If I add the Node manually with MAC address, it shows up and starts commissioning. According to the guide, it should show up atomagically.

Commissioning fails, output from MAAS log

Dec 21 14:27:59 ubuntu maas.import-images: [WARNING] I/O error while syncing boot images. If this problem persists, verify network connectivity and disk usage.
Dec 21 14:27:59 ubuntu maas.import-images: [WARNING] Finished importing boot images, the region does not have any boot images available.
Dec 21 14:30:12 ubuntu maas.node: [INFO] Node-001: Status transition from COMMISSIONING to FAILED_COMMISSIONING
Dec 21 14:30:12 ubuntu maas.node: [ERROR] Node-001: Marking node failed: Node operation 'Commissioning' timed out after 0:20:00.
  • would you mind filing a bug against MAAS? it seems like it should be trivial for maas to know that the nodes in the subnet will not reach it's network without special treating. It should attempt to use the interface that is on the correct subnet. – dpb Dec 21 '15 at 18:47
  • I reported a bug here: bugs.launchpad.net/maas/+bug/1528394 – ladrua Dec 22 '15 at 0:14

I've had a situation like this where I had to configure the MAAS URL correctly. The nodes need to be able to see it. Could you just try:

sudo dpkg-reconfigure maas-region-controller
sudo dpkg-reconfigure maas-cluster-controller

and make sure the URL/IP is the IP reachable by the nodes. After this, try your enlistment again.

For MAAS to be "working" you need to be able to enlist, commission, then deploy each node. It's a good litmus test before you run the autopilot.

| improve this answer | |
  • Praise the lord, and you! That was it. Seems to be working now. Thank you! – ladrua Dec 21 '15 at 14:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.