Re: Fedora 21 Server problems: teamd and NetworkManager :now RHBZ 1172734
by colin
Thanks Jiri
>
> Weird thing is that teamd is started and killed at the very second. I
> suspect NM from doing this. Maybe you have some teamd connection in NM?
>
> What "nmcli con" command says?
I checked 'nmcli connection' and recall no mention of a new team device there.
Although it does correctly report details of a new team device when if
I create it with nmcli.
>
> Weird indeed. Feel free to cc NM people/maling-list.
>
This is a regression from F20, so I opened a NetworkManager bug :
https://bugzilla.redhat.com/show_bug.cgi?id=1172734
9 years, 4 months
- 2
- 2
Fedora 21 Server problems: teamd and NetworkManager
by colin
Hello to all on the list!
I have been using F21Server as a KVM hosted development and test
platform ever since the first Alpha Release and it has been great. :-)
I do appreciate that those of you at Fedora must have been mad busy
these last few weeks and months working to get F21 out the door.
I am using linux containers (lxc) and Open Vswitch (ovs) and have
been learning docker.
My intention was to create a team device and move it into the
namespace of a (docker) container but recently I seem to have hit a
problem.
In searching to reduce the test case to the simplest possible I have found that
'teamd --team-dev teamRR2 -d'
appears to succeed, but when I try to use device teamRR2 it is no
longer there! :-(
I have already sorted out various SELinux permissions, so do not think
that it is that.
/var/log/audit/audit.log shows me no clues.
Excert from /var/log/messages
Dec 8 13:11:59 localhost kernel: [ 4940.990171] IPv6:
ADDRCONF(NETDEV_UP): teamRR2: link is not ready
Dec 8 13:11:59 localhost NetworkManager[682]: <error>
[1418044319.857750] [devices/nm-device-team.c:241]
update_connection(): (teamRR2): failed to read teamd config (err=-110)
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2):
device state change: unmanaged -> unavailable (reason 'managed') [10
20 2]
Dec 8 13:11:59 localhost NetworkManager[682]: <error>
[1418044319.859446] [platform/nm-linux-platform.c:2250] link_change():
Netlink error changing link 18: <UP> mtu 0 (1): No such device
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2):
preparing device
Dec 8 13:11:59 localhost NetworkManager[682]: <info> Deactivation
(teamRR2) stopping teamd...
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2):
device state change: unavailable -> unmanaged (reason 'removed') [20
10 36]
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2): carrier is OFF
Dec 8 13:11:59 localhost kernel: IPv6: ADDRCONF(NETDEV_UP): teamRR2:
link is not ready
Dec 8 13:11:59 localhost NetworkManager: Got SIGINT, SIGQUIT or SIGTERM.
Dec 8 13:11:59 localhost NetworkManager: Exiting...
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2): new
Team device (driver: 'team' ifindex: 19)
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2):
exported as /org/freedesktop/NetworkManager/Devices/17
Dec 8 13:11:59 localhost NetworkManager: Daemon not running
Dec 8 13:11:59 localhost NetworkManager[682]: <info> Activation
(teamRR2) started teamd...
Dec 8 13:11:59 localhost NetworkManager[682]: <error>
[1418044319.890271] [devices/nm-device-team.c:208]
ensure_teamd_connection(): (teamRR2): failed to connect to teamd
(err=-22)
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2):
device state change: unmanaged -> unavailable (reason 'managed') [10
20 2]
Dec 8 13:11:59 localhost NetworkManager[682]: <error>
[1418044319.890801] [platform/nm-linux-platform.c:2250] link_change():
Netlink error changing link 19: <UP> mtu 0 (1): No such device
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2):
preparing device
Dec 8 13:11:59 localhost NetworkManager[682]: <info> Deactivation
(teamRR2) stopping teamd...
Dec 8 13:11:59 localhost NetworkManager[682]: <info> (teamRR2):
device state change: unavailable -> unmanaged (reason 'removed') [20
10 36]
Dec 8 13:11:59 localhost NetworkManager: libteamdctl:
teamdctl_connect: Failed to connect using all CLIs.
If it matters, eth0 is attached to a Linux Bridge
and eth1 attached to an OVS Bridge.
I did apply a large yum update yesterday, including NetworkManager and systemd
but before that NetworkManager would actually crash (leaving the VM
with no networking)
when I tried to create a team device.
Any suggestions or request for further information / tests will be
greatly appreciated and acted upon rapidly.
Thankyou All,
Colin
9 years, 4 months
- 2
- 3