Node keeps status "Deploying"


#1

Hi everyone,

I’m having troubles deploying new servers with MaaS, on a first time I got problems to remove LVM arrays and setting this command : “sudo dd if=/dev/zero of=/dev/sda bs=1M count=400” on each disk in my curtin config solved the problem (thanks to an other post :slight_smile: ).

But now my servers keep status “Deploying” even when it is finished with those messages at the end :

–2019-01-14 14:37:34-- http://192.168.172.3:5240/MAAS/metadata/latest/by-id/k336ee/
Connecting to 192.168.172.3:5240… connected.
HTTP request sent, awaiting response… 200 OK
Length: 2 [text/plain]
Saving to: ‘/dev/null’
0K 100% 202K=0s
2019-01-14 14:37:34 (202 KB/s) - ‘/dev/null’ saved [2/2]
curtin: Installation finished.

Do you have any idea ? Can it be linked with the new command I use to erase LVM arrays ?

Edit :
When it reboots at the end I can see a lot of those messages via console :


I don’t know if it means something because later I can see this :

Thank you !
Ludwig


#2

From what I understand it would mean that DHCP is out of IP.
I did arrive at the end of my pool today but it started to take old adresses no longer used and my server has an IP attributed, I can see it in GUI and ifconfig.
Moreover I have no clue if it is related with my “Deploying” status.


#3

I’m still stuck here and I’m not able to deploy servers anymore, it is not a DHCP problem, it seems my server are not able to connect to curtin datasource on the next reboot after install.
But at the end of the deployment, just before the reboot, I can see it connects well to the database :

Jan 15 20:01:14 dc1-8gpu-77 ec2: -----END SSH HOST KEY FINGERPRINTS-----
Jan 15 20:01:14 dc1-8gpu-77 ec2: #############################################################
Jan 15 20:01:14 dc1-8gpu-77 cloud-init[3397]: Cloud-init v. 18.4-0ubuntu1~16.04.2 running ‘modules:final’ at Tue, 15 Jan 2019 19:57:00 +0000. Up 46.91 seconds.
Jan 15 20:01:14 dc1-8gpu-77 cloud-init[3397]: Cloud-init v. 18.4-0ubuntu1~16.04.2 finished at Tue, 15 Jan 2019 20:01:14 +0000. Datasource DataSourceMAAS [http://192.168.172.3:5240/MAAS/metadata/curtin]. Up 301.31 seconds
Jan 15 20:01:14 dc1-8gpu-77 systemd[1]: Started Execute cloud user/final scripts.
Jan 15 20:01:14 dc1-8gpu-77 systemd[1]: Reached target Cloud-init target.
Jan 15 20:01:14 dc1-8gpu-77 systemd[1]: Startup finished in 13.357s (kernel) + 4min 48.134s (userspace) = 5min 1.491s.
Jan 15 20:01:15 dc1-8gpu-77 systemd[1]: Unmounting /media/root-ro…
Jan 15 20:01:15 dc1-8gpu-77 systemd[1]: Stopping ACPI event daemon…
Jan 15 20:01:15 dc1-8gpu-77 systemd[1]: Closed Load/Save RF Kill Switch Status /dev/rfkill Watch.
Jan 15 20:01:15 dc1-8gpu-77 systemd[1]: Stopping Authenticate and Authorize Users to Run Privileged Tasks…
Jan 15 20:01:15 dc1-8gpu-77 systemd[1]: Stopped target Cloud-init target.

Then it reboots and it displays error messages of my last comment, without being able to connect to cloud-init datasource.
I tried to restart MaaS.

Thank you !
Ludwig


Add 2nd region controller and remove 1st one
#4

Hello,

I figured out what were wrong with my Ubuntu 18 deployment.

In my preseed I was running this late command :

[“curtin”, “in-target”, “–”, “sh”, “-c”, “sudo apt-get -y install vlan ifenslave”]

I need these packages on Ubuntu 16 to make my network working and I reused this preseed file to run 18.04 deployment.
However installing ifenslave package in late commands simply ruins the Netplan config built with the deployment…
My node was not able to communicate with MaaS at the real end of deployment and so it kept “Deploying” status and then it failed.

Ludwig