-
Notifications
You must be signed in to change notification settings - Fork 76
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
"Restart the networking service" failing repeatedly (during IIAB install, and when ./iiab-network is run) [similar Debian 11 Bullseye VM has a strange DNS-like freezing issue] [QA Automation / testing cloud-init Multipass script example] #3568
Comments
The bottommost 10 lines above correspond to:
|
|
EXPLANATION: Debian 11 VM's do not work properly with 2 Ethernet interfaces within Multipass, as demonstrated by this vanilla/fresh "OS ONLY VM" here, prior to installing IIAB...
An "OS ONLY" Debian 11 VM with just 1 Ethernet interface works better, as the paste below shows: (still the above mess suggests Debian 11 networking cannot be trusted within Multipass VM's; which is very unfortunate, as Debian 12 VM's work really well within Multipass, allowing for accelerated testing!)
|
Background: @EMG70 is working on QA Automation of IIAB unit tests using clout-init scripts like this iiab-pbx.yml example...
Then he launches everything (fresh VM + OS install + IIAB install + unit test) thanks to one single command — as follows:
ETC. |
'network_enabled: False' and manage the network on your own. |
Gather the information from a fresh VM without ever installing IIAB, seems like '--bridged --cloud-init ' assembled this config file:
and the resulting network looked like:
Seems like a bit of a bug/misconfiguration somewhere, the above only has ONE default route while the other OS's would show 2 default routes but with different metrics assigned to each. Now without the second gateway present that interface would not be part of 'exclude_devices' via 'second_gateway_found' becoming part of 'lan_list_result' that would become a slave device under br0. Wonder why there is no mention of the devices enp5s0 or enp6s0 in the configuration file. |
I might be wrong, but I'm assuming it's Multipass's So be it. Still, I'll post the 2 things you mention, in case there's something to learn here:
Any others? |
That would be a good start. Think I can gather enough info from ip r |
Think it more of an issue within the pre-canned images not playing nice with more than one network interface. |
Ok. And interestingly, even with "1 single Ethernet" interface, the other Debian 11 (Multipass VM) consistently acted very strangely in a very different way:
|
Think that is a result of the image using the older ifupdown functions where the hostname change becomes effective upon reboot. With a reboot does sudo fail to resolve the hostname of the VM? |
"OS ONLY" Debian 11 VM with 2 Ethernet interfaces are such a mess, they appear to barely reboot, But on first boot (all you get?) it looks like this:
|
"OS ONLY" Debian 11 VM with 1 single Ethernet interface:
|
Apologies that original single-Ethernet-interface ( Just FYI/FWIW it had contained IIAB with roles/pbx successfully installed. Just FYI/FWIW minimal such Multipass VM's with 1 single Ethernet interface have all rebooted and generally (appeared to) work fine prior to installation of IIAB. |
The irony is that Debian 11 (with 2 Ethernet interfaces) works well (as a Multipass VM) until it's rebooted 🤔
RECAP:
|
We can tag this issue as "wontfix" if it's too much trouble. Debian 11 networking is probably patchable after its first boot, to make it work with Multipass & FreePBX etc. But if the effort is too complicated in the end, let's drop it. |
Thanks to @EMG70. Here are some basic diagnostics (from the Debian 11 VM) until we understand more:
iiab-diagnostics: http://sprunge.us/ui5O7W?en
The text was updated successfully, but these errors were encountered: