Wired network problem: "carrier lost" immediately after getting IP

halun · 2009-11-17 13:15:46

My first post here! I hope I'll not commit any serious blunders...

Anyway, here's my problem. I'm running Arch64 on a Lenovo Thinkpad X200s. Usually the wired network connection works flawlessly; just "netcfg wired", where the "wired" config file is a copy of /etc/network.d/examples/ethernet-dhcp. However, at the moment (and for the next two weeks) I'm staying at a hotel which seems to have a really strange network setup. I stayed there last year too, and had all kinds of weird problems with another computer.

What happens at the hotel is that I obtain an IP address with DHCP, and then the connection is lost almost immediately. There is a short time interval (less than a second) where I can, for example, load a web page. (Before I try to connect, I type in the address of a fast-loading site like www.google.com in Firefox, and of course I get an error message about being unable to load the page. Then I start the DHCP client in a terminal window, and while it is trying to connect I keep pressing the reload button in Firefox repeatedly. After a while, the page loads successfully. I never see the login page that's supposed to come up first. But if I try to reload again after a few seconds, there is again an error message.)

Luckily, during the day I'm in an office where the connection works as it should. Otherwise, I don't think I would have been able to post this...

---

Here are the details. First attempt:

[root@fjodor hans]# netcfg wired
:: wired up                                                              [BUSY] dhcpcd: eth0: MTU restored to 1500
dhcpcd: timed out
 - DHCP IP lease attempt failed.

Not very useful. Next, try dhcpcd manually:

[root@fjodor hans]# dhcpcd -B eth0
dhcpcd: version 5.1.3 starting
dhcpcd: eth0: rebinding lease of 192.168.1.71
dhcpcd: eth0: acknowledged 192.168.1.71 from 192.168.1.1 `ÿ'
dhcpcd: eth0: checking for 192.168.1.71
dhcpcd: eth0: carrier lost
dhcpcd: eth0: MTU restored to 1500
dhcpcd: eth0: carrier acquired
dhcpcd: eth0: rebinding lease of 192.168.1.71
dhcpcd: eth0: acknowledged 192.168.1.71 from 192.168.1.1 `ÿ'
dhcpcd: eth0: checking for 192.168.1.71
dhcpcd: eth0: leased 192.168.1.71 for 86400 seconds
dhcpcd: eth0: MTU set to 576
dhcpcd: eth0: carrier lost
dhcpcd: eth0: carrier acquired
dhcpcd: eth0: rebinding lease of 192.168.1.71
^Cdhcpcd: received SIGINT, stopping
dhcpcd: eth0: removing interface

Sometimes the following message also appears in the midst of everything:

cat: /var/run/dhcpcd/resolv.conf.eth0: No such file or directory

The thread http://bbs.archlinux.org/viewtopic.php?id=79923 suggests using dhclient instead, so let's try that:

[root@fjodor hans]# dhclient eth0
Internet Systems Consortium DHCP Client V3.1.2p1
Copyright 2004-2009 Internet Systems Consortium.
All rights reserved.
For info, please visit http://www.isc.org/sw/dhcp/

can't create /var/state/dhcp/dhclient.leases: No such file or directory
Listening on LPF/eth0/00:1f:16:1e:01:ae
Sending on   LPF/eth0/00:1f:16:1e:01:ae
Sending on   Socket/fallback
DHCPDISCOVER on eth0 to 255.255.255.255 port 67 interval 3
DHCPDISCOVER on eth0 to 255.255.255.255 port 67 interval 3
DHCPDISCOVER on eth0 to 255.255.255.255 port 67 interval 6
DHCPOFFER from 192.168.1.1
DHCPREQUEST on eth0 to 255.255.255.255 port 67
DHCPACK from 192.168.1.1
can't create /var/state/dhcp/dhclient.leases: No such file or directory
bound to 192.168.1.71 -- renewal in 36192 seconds.

Well, it looks good, but no connection (except for that half second):

[root@fjodor hans]# ifconfig eth0
eth0      Link encap:Ethernet  HWaddr 00:1F:16:1E:01:AE  
          UP BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:107331 errors:0 dropped:0 overruns:0 frame:0
          TX packets:30444 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:38226834 (36.4 Mb)  TX bytes:4886661 (4.6 Mb)
          Memory:f2600000-f2620000 
[hans@fjodor ~]$ ping 192.168.1.1
connect: Network is unreachable

If I run dhclient in one terminal window, and meanwhile ping repeatedly in another, I get

[hans@fjodor ~]$ ping 192.168.1.1
connect: Network is unreachable
[hans@fjodor ~]$ ping 192.168.1.1
connect: Network is unreachable
[hans@fjodor ~]$ ping 192.168.1.1
connect: Network is unreachable
[hans@fjodor ~]$ ping 192.168.1.1
PING 192.168.1.1 (192.168.1.1) 56(84) bytes of data.
64 bytes from 192.168.1.1: icmp_seq=1 ttl=64 time=6.55 ms
64 bytes from 192.168.1.1: icmp_seq=2 ttl=64 time=2.20 ms
64 bytes from 192.168.1.1: icmp_seq=3 ttl=64 time=2.08 ms
64 bytes from 192.168.1.1: icmp_seq=4 ttl=64 time=4.00 ms
64 bytes from 192.168.1.1: icmp_seq=5 ttl=64 time=2.20 ms
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
ping: sendmsg: Network is unreachable
^C
--- 192.168.1.1 ping statistics ---
13 packets transmitted, 5 received, 61% packet loss, time 12016ms
rtt min/avg/max/mdev = 2.086/3.410/6.551/1.725 ms

By the way, the error message "can't create /var/state/dhcp/dhclient.leases: No such file or directory" above seems strange. Is this a bug? According to the man page, the default file is /var/lib/dhclient/dhclient.leases. If I try "dhclient -lf /var/lib/dhclient/dhclient.leases eth0", then it creates the log file

lease {
  interface "eth0";
  fixed-address 192.168.1.71;
  option subnet-mask 255.255.255.0;
  option routers 192.168.1.1;
  option dhcp-lease-time 86400;
  option dhcp-option-overload 3;
  option dhcp-message-type 5;
  option domain-name-servers 192.168.1.1;
  option dhcp-server-identifier 192.168.1.1;
  option broadcast-address 255.255.255.255;
  renew 2 2009/11/17 07:18:59;
  rebind 2 2009/11/17 18:20:28;
  expire 2 2009/11/17 21:20:28;
}

but the connection is still dropped immediately, just like before.

I talked to the hotel manager, and he was very surprised that I was having trouble, because the network was working on the computer in his office, and apparently also for the other guests.

[I don't know if it's relevant, but the problems I had last year had to do with the login system. The normal procedure was that one had to start a web browser and enter user id and password on a login page before one could start browsing. However, this only applied to http connections. I could use ssh and https even when (as usually happened during the weekends...) the server that was handling the login was down.]

I know next to nothing about what's going on under the hood with all this DHCP stuff, so I have no clue what to try next. Any suggestions?

jcolinzheng · 2009-12-23 08:04:15

See http://bbs.archlinux.org/viewtopic.php?id=86972

halun · 2009-12-24 18:21:28

I'm not able to try it out, since I've left the hotel already. But thanks for the tip anyway!

Arch Linux

#1 2009-11-17 13:15:46

Wired network problem: "carrier lost" immediately after getting IP

#2 2009-12-23 08:04:15

Re: Wired network problem: "carrier lost" immediately after getting IP

#3 2009-12-24 18:21:28

Re: Wired network problem: "carrier lost" immediately after getting IP

Board footer