-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Charles Curley wrote:
> No. Are there user configuration files for NM? That's the only thing
> I can think of that would make this test useful.
I think most of the data on what networks is stored per-user. The
passwords are stored in the gnome-keyring, so that's why I was
thinking it might be good to test as a new user.
> Anyway, I tried it, and got nowhere useful, except I got this:
Bah. Well that sucks.
> [root@dragon ~]# ifconfig
> dev23266 Link encap:Ethernet HWaddr 00:10:C6:C0:AA:26
> UP BROADCAST MULTICAST MTU:1500 Metric:1
> RX packets:0 errors:0 dropped:0 overruns:0 frame:0
> TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:1000
> RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
[...]
> OK, where are the mutant device names coming from? I had created the
> new user, logged out as my usual user, then logged in as the new
> user. I fired up NM as root, of course. Eths 0 and 1 were there
> before and no mutant device names. Oh, and the effort failed.
Ya got me. I don't know enough about debugging NM to help much. :(
>> What's in /var/log/messages? Obviously, NM puts a lot of logging
>> info there, but perhaps if you scan through it you'll find some
>> obvious problem that's just not making it back to you via the GUI.
>
> I don't see anything obvious. Here's a wierd one:
>
> Deactivating device dev3514.
>
> I see other more reasonable device names like eth0 and eth1, but
> that's another mutant device name.
So the devXXXXX numbers change frequently too? If you grep the logs
for one of those, are there any other entries for them which might
give a clue as to when or why they're being created?
(And yeah, I'm just taking stabs here.)
> I see quite a few of these:
>
> Activation (eth0/wireless): association took too long (>20s),
> failing activation.
Those don't bode well for a good connection. I've had NM drop my
connections randomly (always when I'm in a hurry and trying to do
something that hits the network, like opening an image folder on an
NFS share in nautilus). And I usually see lines similar to that in my
logs afterward. I've taken to restarting NM when I see that, as
otherwise it takes a few minutes before it will pick up my home
network and reconnect itself.
> Since the keys haven't changed and I can log in with "service
> network start", I suspect a bug in NM. But then again, I get similar
> messages using a kernel and version of NM that used to work. Ideas?
It might be worth asking about this on the NM list. Hopefully some of
the folks there will be able to give better pointers on how to debug
this.
> The "activation" lines are interlaced with a slew of WPA supplicant
> lines. Since WPA is not an issue on this network, I wonder if WPA
> Supplicant and NM are tripping over each other. However, I looked at
> some old logs that show sucessful activations. They show similar
> interlacing, so I don't think that's the issue.
I don't think so either. The wpa_supplicant control interface is just
very verbose (it's patched in the Fedora packages to be that way), and
NM uses interface.
- --
Todd OpenPGP -> KeyID: 0xD654075A | URL: www.pobox.com/~tmz/pgp
======================================================================
I think it would be a good idea.
-- Mohandas Gandhi (1869-1948),
when asked what he thought of Western civilization
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.4 (GNU/Linux)
Comment: When crypto is outlawed bayl bhgynjf jvyy unir cevinpl.
iG0EARECAC0FAkSwEjkmGGh0dHA6Ly93d3cucG9ib3guY29tL350bXovcGdwL3Rt
ei5hc2MACgkQuv+09NZUB1o+VgCg45CJ8o11nRMyZsPC0cQ3qa3m7B0AoO3WCoR3
DaSVmxZ8HLE1J7u8xxUt
=j4di
-----END PGP SIGNATURE-----
