NetworkManager doesn't work if latest dhcp3-client package is installed

Bug #32223 reported by Celso Pinto
38
Affects Status Importance Assigned to Milestone
dhcdbd (Ubuntu)
Fix Released
Medium
Martin Pitt

Bug Description

I upgraded my Dapper system and this upgrade included updates to the dhcp3-client and dhcp3-common packages. After rebooting NetworkManager stopped working. It connected to the AP but never got an IP address. The log showed this message:

Feb 21 02:10:34 localhost NetworkManager: <information>^IActivation (eth1) Stage 3 (IP Configure Start) started...
Feb 21 02:10:35 localhost NetworkManager: <information>^IActivation (eth1) Beginning DHCP transaction.
Feb 21 02:10:35 localhost dhclient: Internet Systems Consortium DHCP Client V3.0.3
Feb 21 02:10:35 localhost dhclient: Copyright 2004-2005 Internet Systems Consortium.
Feb 21 02:10:35 localhost dhclient: All rights reserved.
Feb 21 02:10:35 localhost dhclient: For info, please visit http://www.isc.org/products/DHCP
Feb 21 02:10:35 localhost dhclient: Usage: dhclient [-1dqr] [-nw] [-p <port>] [-s server]
Feb 21 02:10:35 localhost dhclient: [-cf config-file] [-lf lease-file][-pf pid-file] [-e VAR=val]
Feb 21 02:10:35 localhost dhclient: [-sf script-file] [interface]

This occurs because NetworkManager invokes dhclient with the following arguments:
-r -lf /var/lib/dhcp3/dhclient.eth1.leases -pf /var/run/dhclient.eth1.pid -q -e dhc_dbus=31 -x -d eth1

And -x is no longer a valid argument.

The only solution possible is to revert to an older version.

Revision history for this message
Aneesh Kumar K.V (aneesh-kumar) wrote :

Dropping the -x option doesn't help. I tried with a locally build dhcdbd package. Network-manager enters into an zeroconf setup where in it gets a useless IP.

Feb 21 12:39:07 localhost ntpdate[5495]: no server suitable for synchronization found
Feb 21 12:39:21 localhost NetworkManager: <information>^IDevice 'eth1' DHCP transaction took too long (>25s), stopping it.
Feb 21 12:39:21 localhost NetworkManager: <information>^IActivation (eth1) Stage 4 (IP Configure Timeout) scheduled...
Feb 21 12:39:21 localhost NetworkManager: <information>^IActivation (eth1) Stage 4 (IP Configure Timeout) started...
Feb 21 12:39:21 localhost NetworkManager: <information>^INo DHCP reply received. Automatically obtaining IP via Zeroconf.
Feb 21 12:39:21 localhost NetworkManager: <information>^Iautoip: Sending probe #0 for IP address 169.254.244.174.

The only way to get around this is by disabling eth1 in /etc/network/interfaces and using ifup eth1 after the system booted.

Revision history for this message
Christian Kellner (gicmo) wrote :

Confirming this. It never worked for me with my wirless card but now it even stopped for the wired network.

Revision history for this message
Martin Pitt (pitti) wrote :

Found the bug, fixing now.

Changed in network-manager:
assignee: nobody → pitti
status: Unconfirmed → Confirmed
Revision history for this message
Martin Pitt (pitti) wrote :

 dhcdbd (1.10-0ubuntu11) dapper; urgency=low
 .
   * debian/rules: Disable dhclient extended options (-x) since the current
     version doesn't have it any more (our NetworkManager doesn't need it).
     (Malone #32223)

 dhcp3 (3.0.3-6ubuntu6) dapper; urgency=low
 .
   * debian/dhclient-script.linux: dhddbd needs to be told about the
      result of the client script; so instead of notifying it at the beginning
      and immediately exiting the script, call dbus-send to dhcdbd in the
      exit_with_hooks() function, so that resolv.conf is written properly.
      (Malone #32134, #32223)

Changed in dhcdbd:
status: Confirmed → Fix Released
Revision history for this message
fangorious (fangorious-deactivatedaccount) wrote :

Coming in from bug 32134 today's update fixed it for me.

Revision history for this message
Tristan Wibberley (tristan-wibberley) wrote :

My installation is up to date but I'm still seeing something strange going on...

$ ifconfig eth0
eth0 Link encap:Ethernet HWaddr 00:50:8D:EF:D8:CB
          inet addr:169.254.37.240 Bcast:0.0.0.0 Mask:255.255.0.0
          inet6 addr: fe80::250:8dff:feef:d8cb/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
          RX packets:3744 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2150 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:5075408 (4.8 MiB) TX bytes:175752 (171.6 KiB)
          Interrupt:193 Base address:0xb400

$ route -n
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth0
192.168.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth0
0.0.0.0 192.168.1.1 0.0.0.0 UG 0 0 0 eth0

$ cat /etc/network/interfaces
# This file describes the network interfaces available on your system
# and how to activate them. For more information, see interfaces(5).

# The loopback network interface
auto lo
iface lo inet loopback

# This is a list of hotpluggable network interfaces.
# They will be activated automatically by the hotplug subsystem.
auto eth0

# The primary network interface
#iface eth0 inet dhcp
iface eth0 inet static
        address 192.168.1.35
        gateway 192.168.1.1
        netmask 255.255.0.0

Revision history for this message
Scott James Remnant (Canonical) (canonical-scott) wrote :

This isn't caused by NetworkManager, but by a program such as "zeroconf" if you have that installed.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.