Comment 23 for bug 184711

Revision history for this message
eryksun (eryksun) wrote : Re: [Bug 184711] Re: network-admin error: The interface does not exist

Did you try to run 'network-admin' directly as a non-root user without any
options?

* If I run as root with 'gksudo network-admin', the window comes up grayed
out; I can't unlock it; and there's an error reported in the CLI: "Unable
to lookup session information for process" (IMO, the GtkComboBox error is
irrelevant).

* If I run 'network-admin -c eth0', it goes straight to a message box saying
"The interface does not exist...", but there's no "Unable to lookup session
information..." error in the CLI.

* If I run " gksudo 'network-admin -c eth0' ", I get both errors, and the
message box window for "The interface does not exist" is using the GTK theme
of the root user, which is very different from my user theme on purpose to
know when a window is running as root.

* When I click the 'configure' button in gnome-nettool, I get the same two
errors on the CLI, and the error message is also a root window. I think
gnome-nettool is simply calling 'sudo network-admin -c INTERFACE', which no
longer seems to work -- doubly.

* If I run just 'network-admin', the window loads normally; I'm able to
unlock it, select an interface and configure it.

On Tue, Jul 8, 2008 at 8:31 AM, Filofel <email address hidden> wrote:

> Variant here:
> Running Hardy.
>
> I needed to change from 100MBps to 1000MBps.
> Shutdown, replaced the Intel Pro/100 NIC (removed it) for a Pro/1000, fixed
> the DHCP so the machine kept the same IP.
> At reboot, the machine had detected the new NIC and assigned it to eth1.
> The network worked.
> But the "Configure" button in Networks Tools behaved as reported here.
> Incidentally, /etc/network/interfaces was still referencing eth0. Changing
> it to eth1 didn't make anything better (even after a reboot) - at least
> anything obvious.
>
> I manually put eth0 back in interfaces, shutdown, removed the Pro/1000 for
> the original e100b NIC, fixed the DHCP server accordingly: The machine went
> back to using eth0. But the problem persisted (same "The interface does not
> exist" message.
> So it looks like I'm just in the same situation as anyone else here. I
> might just have got there using a slightly different path.
>