[SOLVED] DHCP-Issue since last update

rayquaza · 2014-02-18 13:39:21

Yesterday I did 'pacman -Syu' and rebooted. Since then I can't use the network-interface. journalctl reports

Job sys-subsystem-net-devices-multiuser.device/start timed out
Timed out waiting for device sys-subsystem-net-devices-multiuser.device
Dependency failed for dhcpcd on multi/user

and later the following every minute (where 192.168.1.1 is my router):

enp2s0: NAK: via 192.168.1.1

Somewhere on this board I found the hint to comment-out "require dhcp_server_identifier" in /etc/dhcpcd.conf, which doesn't work for me. After assigning a static IP with "ifconfig enp2s0 down; ifconfig enp2s0 192.168.1.66; ifconfig enp2s0 up" I can ping from and to the system using IPs, but not it's hostname and name-resolving doesn't work.
Any ideas on how to get it working properly again?

Last edited by rayquaza (2014-02-20 10:56:03)

fdservices · 2014-02-18 15:56:35

Can you post the output of

# ip addr

Andrew

rayquaza · 2014-02-18 16:08:18

Sure: (scrambled MAC and IPv6)

$ ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: enp2s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 00:1d:91:2a:7d:a6 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.66/24 brd 192.168.1.255 scope global enp2s0
       valid_lft forever preferred_lft forever
    inet6 fe80::21d:36ff:fea3:c69a/64 scope link 
       valid_lft forever preferred_lft forever
3: wlp3s0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN group default qlen 1000
    link/ether 00:22:62:74:f8:13 brd ff:ff:ff:ff:ff:ff

fdservices · 2014-02-18 16:29:46

OK, enp2s0 is connected, link is up and you have an address.

I take it that this was before your "fix"?

Have you populated /etc/hosts?

Check to routing

# route -n

Andrew

rayquaza · 2014-02-18 16:56:29

fdservices wrote:

I take it that this was before your "fix"?

Ouch, right… Of course this was after assigning a static IP. Output after a reboot (and finding a portable disk):

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default 
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: enp2s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
    link/ether 00:1d:91:2a:7d:a6 brd ff:ff:ff:ff:ff:ff
    inet6 fe80::21d:36ff:fea3:c69a/64 scope link 
       valid_lft forever preferred_lft forever
3: wlp3s0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc mq state DOWN group default qlen 1000
    link/ether 00:22:62:74:f8:13 brd ff:ff:ff:ff:ff:ff

fdservices wrote:

Have you populated /etc/hosts?

Only what the Install-Guide instructs (nothing?), so it's content is

#<ip-address>   <hostname.domain.org>   <hostname>
127.0.0.1       localhost.localdomain   localhost
::1             localhost.localdomain   localhost

# End of file

$ route -n
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface

fdservices · 2014-02-19 08:13:19

OK, enp2s0 has no IPv4 address so you cannot get anywhere.

Is dhcpcd running?

 $ ps -el | grep dhc

Look for the line "inet 192.168.1.66/24 brd 192.168.1.255 scope global enp2s0"

I sometimes find that dhclient works better for me so you could try that instead. If you want to use it then install it and

# dhclient -v enp2s0

You should be able to assign static ip addresses on your router for each destination on your network otherwise check https://wiki.archlinux.org/index.php/Ne … IP_address

Then you can add your network destinations to the /etc/hosts file so that the names are resolved to the addresses of the destinations e.g.

127.0.0.1 My_Hostname localhost
::1       localhost
192.168.1.1 My_Router
192.168.1.2 Another_PC 
192.168.1.3 My_Network_Printerl
etc

Let us know how you get on

Andrew

rayquaza · 2014-02-19 12:16:10

dhcpcd is running (how else could it print to syslog every minute?). From how I understand wireshark's output the router is offering my normal IP, my system is requesting this (same) IP and the router decides to say no‽
I even tried downgrading dhcpcd, which didn't help. I just realized my network admin was talking about a router-update a few days ago (my system wasn't off since then, only in suspend-to-ram), so I'll ask him about a downgrade for testing.

In case I'd want to downgrade my whole system to a specific day: I would have to edit my /etc/pacman.d/mirrorlist as written here (and get the connection to that server working) and do a pacman -Syyu, right?

fdservices · 2014-02-19 14:54:51

Dependency failed for dhcpcd on multi/user

May be you should reboot then to make sure the updates have been correctly applied?

Andrew

rayquaza · 2014-02-19 15:07:56

Oh, by "wasn't off since then" I meant until the update I rebooted after every change which might potentially need it (=after every change mentioned in this thread).

fdservices · 2014-02-19 15:10:17

Reboot the router?

Andrew

Axel Svensson · 2014-02-19 18:40:26

Got similar behavior on my system today after update and reboot. Could not ping any external IP. Turned out in the end that it was because of firewall settings in iptables. I guess iptables have been updated the last couple of days and my firewall rules was not OK. No problem with network after I disabled iptables.

rayquaza · 2014-02-20 09:12:05

Got some good news (well, at least for ArchLinux that is):

I tried downgrading my system as well as updating it further and it didn't help, I tried disabling iptables as suggested and it didn't help. But I got the chance to test it with a different router and it works there, so I guess it's not ArchLinux' fault. Thanks for your attempts anyway.

In case anyone is interested: The router where it doesn't work is a AVM Fritz!Box 7270.

If someone is looking for how to configure it manually: Look at the manpages of ifconfig (for setting a static IP), route (setting the default gateway) and resolv.conf (for setting the nameserver).

fdservices · 2014-02-20 09:51:17

Reboot the router?

Should have just said "Boot the router"

Glad it worked out

Andrew

fdservices · 2014-02-20 09:51:54

Can you mark this thread [SOLVED]?

Andrew

mir · 2014-02-21 22:38:58

I experienced the same. No need to boot or reboot the router - remove the permanent IP allocation on the fritzbox, then restart dhcpcd, then set the permanent IP again.

Background: dhcpcd changed the duid which is transmitted to the dhcpcd server, and when the fritzbox sees a request for a permanently allocated IP and the duids do not match, it NAKs, which happend again and again. I'm not sure if the fault is with dhcpcd or fritzbox.

rayquaza · 2014-02-22 11:31:34

Ah thanks, that worked! I would have expected to get a new IP in this case, but don't know what the specification says.

mir · 2014-03-30 19:22:26

I continued to have annoying issues with dhcp, such as frequent connection loss with need to restart dhcpcd. Removing

option rapid_commit

from /etc/dhcpcd.conf seems to help.

fdservices · 2014-03-31 07:32:03

Did you try dhclient?

Andrew

mir · 2014-04-08 20:32:44

I use wicd, and while it states that it could use either dhclient or dhcpcd, its package requires dhcpcd. Sigh, I still have issues, but I currently don't have the time to sort it out and have resorted to a fixed IP address for the wired network.

fdservices · 2014-04-09 16:55:38

I assume you are using a wired connection, and cannot obtain an address automatically.

Your comment does not tie in with the Wicd - ArchWiki viz:

Wicd skips obtaining IP address on wlp
This can be caused by dhcpcd running alongside wicd as systemd service. A solution would be to stop/disable dhcpcd.
dhcpcd not running
Normally it should not be required, nor recommended to run the dhcpcd service next to wicd. However, if you encounter the error message that dhcpcd is not running, then you can try running dhcpcd and see if you encounter any incompatibilities when using both services at the same time.
# systemctl start dhcpcd
Alternatively, as a workaround you might consider switching to dhclient in the Wicd settings.

Or is it that the connection is being dropped from time to time?

Andrew

mir · 2014-05-05 08:49:01

No, that's not my problem. I get my address, everything works fine, and after some time the connection drops without obvious reason and usually reconnects with some interactions between wicd and dhcpcd I fail to understand completely. The host is a Dell laptop in its docking station,

May  5 10:32:52 localhost dhcpcd[1855]: eno1: carrier lost
May  5 10:32:52 localhost kernel: [36610.329091] e1000e: eno1 NIC Link is Down
May  5 10:32:52 localhost dhcpcd[1855]: eno1: deleting route to 192.168.178.0/24
May  5 10:32:52 localhost dhcpcd[1855]: eno1: deleting default route via 192.168.178.1
May  5 10:32:57 localhost wicd[1263]: SIOCSIFFLAGS: Operation not possible due to RF-kill
May  5 10:32:57 localhost wicd[1263]: wlp3s0    Interface doesn't support scanning : Network is down
May  5 10:32:57 localhost wicd[1263]: dhcpcd[23391]: sending commands to master dhcpcd process
May  5 10:32:57 localhost dhcpcd[23391]: sending commands to master dhcpcd process
May  5 10:32:57 localhost dhcpcd[1855]: control command: /usr/sbin/dhcpcd -k wlp3s0
May  5 10:32:57 localhost wicd[1263]: SIOCSIFFLAGS: Operation not possible due to RF-kill
May  5 10:32:57 localhost wicd[1263]: SIOCSIFFLAGS: Operation not possible due to RF-kill
May  5 10:32:57 localhost wicd[1263]: Failed to connect to non-global ctrl_ifname: wlp3s0  error: No such file or directory
May  5 10:32:57 localhost dhcpcd[23397]: sending commands to master dhcpcd process
May  5 10:32:57 localhost dhcpcd[1855]: control command: /usr/sbin/dhcpcd -k eno1
May  5 10:32:57 localhost dhcpcd[1855]: eno1: removing interface
May  5 10:32:57 localhost wicd[1263]: dhcpcd[23397]: sending commands to master dhcpcd process
May  5 10:32:57 localhost kernel: [36614.732105] e1000e 0000:00:19.0: irq 46 for MSI/MSI-X
May  5 10:32:57 localhost kernel: [36614.833398] e1000e 0000:00:19.0: irq 46 for MSI/MSI-X
May  5 10:32:57 localhost kernel: [36614.833741] IPv6: ADDRCONF(NETDEV_UP): eno1: link is not ready
May  5 10:32:57 localhost wicd[1263]: Failed to connect to non-global ctrl_ifname: eno1  error: No such file or directory
May  5 10:32:57 localhost dhcpcd[1855]: eno1: waiting for carrier
May  5 10:32:58 localhost kernel: [36615.524788] dell_wmi: Received unknown WMI event (0x0)
May  5 10:32:58 localhost logger: ACPI: PNP0C14:00 000000d0 00000000
May  5 10:32:58 localhost logger: ACPI group/action undefined: PNP0C14:00 / 000000d0 / 00000000 /
May  5 10:32:58 localhost ntpd[1297]: Deleting interface #26 eno1, fe80::5e26:aff:fe44:680e#123, interface stats: received=0, sent=0, dropped=0, active_time=1488 secs
May  5 10:32:58 localhost ntpd[1297]: peers refreshed
May  5 10:33:02 localhost wicd[1263]: SIOCSIFFLAGS: Operation not possible due to RF-kill
May  5 10:33:02 localhost wicd[1263]: wlp3s0    Interface doesn't support scanning : Network is down
May  5 10:33:07 localhost wicd[1263]: SIOCSIFFLAGS: Operation not possible due to RF-kill
May  5 10:33:07 localhost wicd[1263]: wlp3s0    Interface doesn't support scanning : Network is down
May  5 10:33:07 localhost dhcpcd[1855]: eno1: carrier acquired
May  5 10:33:07 localhost kernel: [36625.154435] e1000e: eno1 NIC Link is Up 100 Mbps Full Duplex, Flow Control: Rx/Tx
May  5 10:33:07 localhost kernel: [36625.154448] e1000e 0000:00:19.0 eno1: 10/100 speed: disabling TSO
May  5 10:33:07 localhost kernel: [36625.154495] IPv6: ADDRCONF(NETDEV_CHANGE): eno1: link becomes ready
May  5 10:33:07 localhost dhcpcd[1855]: eno1: IAID 0a:44:68:0e
May  5 10:33:07 localhost dhcpcd[1855]: eno1: soliciting a DHCP lease
May  5 10:33:07 localhost dhcpcd[1855]: eno1: offered 192.168.178.52 from 192.168.178.1

Sometimes wicd/dhcpcd failed to reconnect, but it did not happen again after my recent update.

I find this difficult to debug. It might be caused by the network switch, cabling or the dock ...

Arch Linux

#1 2014-02-18 13:39:21

[SOLVED] DHCP-Issue since last update

#2 2014-02-18 15:56:35

Re: [SOLVED] DHCP-Issue since last update

#3 2014-02-18 16:08:18

Re: [SOLVED] DHCP-Issue since last update

#4 2014-02-18 16:29:46

Re: [SOLVED] DHCP-Issue since last update

#5 2014-02-18 16:56:29

Re: [SOLVED] DHCP-Issue since last update

#6 2014-02-19 08:13:19

Re: [SOLVED] DHCP-Issue since last update

#7 2014-02-19 12:16:10

Re: [SOLVED] DHCP-Issue since last update

#8 2014-02-19 14:54:51

Re: [SOLVED] DHCP-Issue since last update

#9 2014-02-19 15:07:56

Re: [SOLVED] DHCP-Issue since last update

#10 2014-02-19 15:10:17

Re: [SOLVED] DHCP-Issue since last update

#11 2014-02-19 18:40:26

Re: [SOLVED] DHCP-Issue since last update

#12 2014-02-20 09:12:05

Re: [SOLVED] DHCP-Issue since last update

#13 2014-02-20 09:51:17

Re: [SOLVED] DHCP-Issue since last update

#14 2014-02-20 09:51:54

Re: [SOLVED] DHCP-Issue since last update

#15 2014-02-21 22:38:58

Re: [SOLVED] DHCP-Issue since last update

#16 2014-02-22 11:31:34

Re: [SOLVED] DHCP-Issue since last update

#17 2014-03-30 19:22:26

Re: [SOLVED] DHCP-Issue since last update

#18 2014-03-31 07:32:03

Re: [SOLVED] DHCP-Issue since last update

#19 2014-04-08 20:32:44

Re: [SOLVED] DHCP-Issue since last update

#20 2014-04-09 16:55:38

Re: [SOLVED] DHCP-Issue since last update

#21 2014-05-05 08:49:01

Re: [SOLVED] DHCP-Issue since last update

Board footer