[opensuse-factory] DHCP broke in Tumbleweed
Since an updated in the last week or so, DHCPv4 doesn't work anymore on my Tumbleweed system. It is connected via ethernet to a router. Neither wickedd nor a manual dhclient can get an IP address. If I configure the IP manually the network works fine. It looks like dhclient is ignoring the replies from the server. Strangely the last dhclient update was a long time ago. dhcp-client-4.2.6-13.1.x86_64 I also tried an older kernel, and killing off avahi, but it didn't make any difference. Any known issues in this area? -Andi 09:34:54.924230 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:34:54.938194 IP 192.168.0.1.67 > 192.168.0.64.68: BOOTP/DHCP, Reply, length 274 09:34:54.947325 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:34:54.947352 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:34:54.955249 ARP, Request who-has 192.168.0.65 tell 192.168.0.1, length 46 09:34:56.983115 IP 192.168.0.1.67 > 192.168.0.65.68: BOOTP/DHCP, Reply, length 274 09:34:59.985178 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:34:59.998000 IP 192.168.0.1.67 > 192.168.0.65.68: BOOTP/DHCP, Reply, length 274 09:35:00.007060 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:35:00.007097 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:35:00.015056 ARP, Request who-has 192.168.0.66 tell 192.168.0.1, length 46 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 2015-05-31 18:42:44 +0200, Andi Kleen wrote:
Since an updated in the last week or so, DHCPv4 doesn't work anymore on my Tumbleweed system. It is connected via ethernet to a router. Neither wickedd nor a manual dhclient can get an IP address. If I configure the IP manually the network works fine.
It looks like dhclient is ignoring the replies from the server. Strangely the last dhclient update was a long time ago.
dhcp-client-4.2.6-13.1.x86_64
I also tried an older kernel, and killing off avahi, but it didn't make any difference.
Any known issues in this area?
Firewall maybe? darix -- openSUSE - SUSE Linux is my linux openSUSE is good for you www.opensuse.org -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 05/31/2015, 06:42 PM, Andi Kleen wrote:
09:34:54.924230 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:34:54.938194 IP 192.168.0.1.67 > 192.168.0.64.68: BOOTP/DHCP, Reply, length 274
Shouldn't the reply be 192.168.0.1.67 > 255.255.255.255.68 ? Didn't your AP/router get crazy? Try to restart it... thanks, -- js suse labs -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 06/01/2015, 02:07 PM, Jiri Slaby wrote:
On 05/31/2015, 06:42 PM, Andi Kleen wrote:
09:34:54.924230 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:34:54.938194 IP 192.168.0.1.67 > 192.168.0.64.68: BOOTP/DHCP, Reply, length 274
Shouldn't the reply be 192.168.0.1.67 > 255.255.255.255.68 ? Didn't your AP/router get crazy? Try to restart it...
In any way, syslog should show what was sent (DISCOVER/REQUEST) and received (OFFER/NAK). Any traces of that? -- js suse labs -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Mon, Jun 01, 2015 at 02:12:56PM +0200, Jiri Slaby wrote:
On 06/01/2015, 02:07 PM, Jiri Slaby wrote:
On 05/31/2015, 06:42 PM, Andi Kleen wrote:
09:34:54.924230 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:34:54.938194 IP 192.168.0.1.67 > 192.168.0.64.68: BOOTP/DHCP, Reply, length 274
Shouldn't the reply be 192.168.0.1.67 > 255.255.255.255.68 ?> Didn't your AP/router get crazy? Try to restart it...
Did not help.
In any way, syslog should show what was sent (DISCOVER/REQUEST) and received (OFFER/NAK). Any traces of that?
I don't see anything on dhcp in the journalctl output. I will debug further. -Andi -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On 06/01/2015 08:07 AM, Jiri Slaby wrote:
On 05/31/2015, 06:42 PM, Andi Kleen wrote:
09:34:54.924230 IP 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from 80:ee:73:16:dc:43, length 300 09:34:54.938194 IP 192.168.0.1.67 > 192.168.0.64.68: BOOTP/DHCP, Reply, length 274 Shouldn't the reply be 192.168.0.1.67 > 255.255.255.255.68 ? Didn't your AP/router get crazy? Try to restart it...
No, the DHCP server responds to the assigned IP address, not the broadcast address. However, it's the MAC address that the client listens for, as with any IP transaction. Also, until the client obtains it's address, it uses 0.0.0.0 as the source address. Are the MAC addresses correct for the client? The server should be sending replies to the client MAC address. BTW, when adding a port number to the end of an IP address, it helps to use a different character, such as ":" to separate it from the IP address. With a ".", it looks like part of the address. So, use 0.0.0.0:68, instead of 0.0.0.0.68. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Sun, 31 May 2015 18:42:44 +0200 Andi Kleen <andi@firstfloor.org> wrote:
Since an updated in the last week or so, DHCPv4 doesn't work anymore on my Tumbleweed system. It is connected via ethernet to a router. Neither wickedd nor a manual dhclient can get an IP address. If I configure the IP manually the network works fine.
When you have wicked installed, could you inspect/provide an output of the following command: sudo /usr/lib/wicked/bin/wickedd-dhcp4 --log-level debug --log-target stderr --debug all --test $IFNAME -- Best Regards, Pawel Wieczorkiewicz <pwieczorkiewicz@suse.de>, Linux System Developer SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Dilip Upmanyu, Graham Norton, HRB 21284 (AG Nürnberg) Maxfeldstraße 5 / 90409 Nürnberg / Germany / Phone: +49-911-740 53 - 613 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Mon, Jun 01, 2015 at 02:23:17PM +0200, Pawel Wieczorkiewicz wrote:
On Sun, 31 May 2015 18:42:44 +0200 Andi Kleen <andi@firstfloor.org> wrote:
Since an updated in the last week or so, DHCPv4 doesn't work anymore on my Tumbleweed system. It is connected via ethernet to a router. Neither wickedd nor a manual dhclient can get an IP address. If I configure the IP manually the network works fine.
When you have wicked installed, could you inspect/provide an output of the following command:
sudo /usr/lib/wicked/bin/wickedd-dhcp4 --log-level debug --log-target stderr --debug all --test $IFNAME
This gets an IP address. Still nothing at boot or ever getting configured. Also restarting the router didn't help. ::: Reading config file /etc/wicked/common.xml ::: subprocess 1830 (/sbin/modprobe -qs -- af_packet) exited with status 0 ::: enp4s0: Link is not up, trying to bring it up ::: ni_system_interface_link_change(enp4s0) ::: bringing up enp4s0 ::: __ni_rtnl_link_up(enp4s0, idx=2) ::: Full refresh of all interfaces (enforced) ::: enp4s0: Link is not (yet) up ::: Full refresh of all interfaces (enforced) ::: Received request: ::: acquire-timeout 10 ::: lease-time 4294967295 ::: start-delay 0 ::: hostname <none> ::: vendor-class <none> ::: client-id 01:80:ee:73:16:dc:43 ::: uuid 115a6c55-6273-0700-2507-000001000000 ::: update-flags arp, hostname, dns, nis, ntp, csr, gateway, mtu ::: recover_lease false ::: release_lease false Notice: enp4s0: Request to acquire DHCPv4 lease with UUID 115a6c55-6273-0700-2507-000001000000 ::: timeout 0 adjusted by 399 to 399 (jr 500) Info: enp4s0: Initiating DHCPv4 discovery (ifindex 2) ::: valid lease: 0; have prefs: 0 ::: enp4s0: setting fsm timeout to 10000 msec ::: sending DHCP4_DISCOVER with xid 0x188e31b ::: timeout 4000 adjusted by 28 to 4028 (jr 2000) ::: arming retransmit timer (4028 msec) ::: enp4s0: incoming packet ::: enp4s0: received DHCP4_OFFER message in state SELECTING Info: enp4s0: Received offer for 192.168.0.6 from 192.168.0.1 ::: dhcp4_tester_protocol_event(ev=0, dev=enp4s0[2], config-uuid=115a6c55-6273-0700-2507-000001000000) INTERFACE='enp4s0' TYPE='dhcp' FAMILY='ipv4' UUID='115a6c55-6273-0700-2507-000001000000' IPADDR='192.168.0.6/24' NETMASK='255.255.255.0' NETWORK='192.168.0.0' BROADCAST='192.168.0.255' PREFIXLEN='24' GATEWAYS='192.168.0.1' DNSSERVERS='192.168.0.1' CLIENTID='01:80:ee:73:16:dc:43' SERVERID='192.168.0.1' ACQUIRED='1433164309' LEASETIME='86400' ::: no sockets left to watch ::: enp4s0: Deleting dhcp4 device with index 2 -- ak@linux.intel.com -- Speaking for myself only. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
On Mon, 1 Jun 2015 15:16:54 +0200 Andi Kleen <andi@firstfloor.org> wrote:
::: enp4s0: received DHCP4_OFFER message in state SELECTING Info: enp4s0: Received offer for 192.168.0.6 from 192.168.0.1 ::: dhcp4_tester_protocol_event(ev=0, dev=enp4s0[2], config-uuid=115a6c55-6273-0700-2507-000001000000) INTERFACE='enp4s0' TYPE='dhcp' FAMILY='ipv4' UUID='115a6c55-6273-0700-2507-000001000000' IPADDR='192.168.0.6/24' NETMASK='255.255.255.0' NETWORK='192.168.0.0' BROADCAST='192.168.0.255' PREFIXLEN='24' GATEWAYS='192.168.0.1' DNSSERVERS='192.168.0.1' CLIENTID='01:80:ee:73:16:dc:43' SERVERID='192.168.0.1' ACQUIRED='1433164309' LEASETIME='86400' ::: no sockets left to watch ::: enp4s0: Deleting dhcp4 device with index 2
This one looks good. So your DHCP appears to be working fine. The --test variant is not supposed to perform full configuration. Just display what it gets. Does configuration work for you when all instances of dhclient are killed and NetworkManager is stopped (systemctl stop NetworkManager) and you perform: wicked --log-target stderr --log-level debug --debug all ifup $ifname given that /etc/sysconfig/network/ifcfg-$ifname contains: BOOTPROTO=dhcp STARTMODE=auto -- Best Regards, Pawel Wieczorkiewicz <pwieczorkiewicz@suse.de>, Linux System Developer SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Dilip Upmanyu, Graham Norton, HRB 21284 (AG Nürnberg) Maxfeldstraße 5 / 90409 Nürnberg / Germany / Phone: +49-911-740 53 - 613 -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
Does configuration work for you when all instances of dhclient are killed and NetworkManager is stopped (systemctl stop NetworkManager) and you perform:
wicked --log-target stderr --log-level debug --debug all ifup $ifname
given that /etc/sysconfig/network/ifcfg-$ifname contains: BOOTPROTO=dhcp STARTMODE=auto
Yes, after I created the ifcfg file (I only had one for eth0 there) and removed NetworkManager the DHCP works again. Thanks! -Andi -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
participants (5)
-
Andi Kleen -
James Knott -
Jiri Slaby -
Marcus Rueckert -
Pawel Wieczorkiewicz