View previous topic :: View next topic |
Author |
Message |
jorro n00b
Joined: 09 Feb 2010 Posts: 20
|
Posted: Mon Mar 10, 2014 12:17 pm Post subject: Network stopped working |
|
|
Hi everyone,
I have a problem similar to this here: https://forums.gentoo.org/viewtopic-t-985024-highlight-.html , even the same network card.
My network does not seem to work anymore since Friday, without any obvious reason. I do not get an IP via dhcp, if I set it to fixed IP, I cannot access anything. The logs pretty much look similar to the above post, with dhcp, I get this:
Code: |
eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 169.254.247.3 netmask 255.255.0.0 broadcast 169.254.255.255
ether <MY_MAC_ADDRESS> txqueuelen 1000 (Ethernet)
RX packets 39373563 bytes 21067758733 (19.6 GiB)
RX errors 0 dropped 1470 overruns 0 frame 0
TX packets 18110079 bytes 5336028576 (4.9 GiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
device interrupt 17
|
When I set it to the fixed IP, I get the correct output with my IP, netmask and broadcast. It was set to fixed IP before it stopped working. dmesg -w returns the same as in the post (I copy it here from the post as I don't want to typo everything):
Code: |
[TIME_STAMP] tg3 0000:05:00.0 eth0: Link is up at 100 Mbps, full duplex
[TIME_STAMP] tg3 0000:05:00.0 eth0: Flow control is off for TX and off for RX
[TIME_STAMP] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[TIME_STAMP] tg3 0000:05:00.0: irq 68 for MSI/MSI-X
[TIME_STAMP] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
|
In the post above an update of udev solved the issues, but I use udev-210 for some time now, everything worked fine.
Any ideas? |
|
Back to top |
|
|
krinn Watchman
Joined: 02 May 2003 Posts: 7470
|
Posted: Mon Mar 10, 2014 9:45 pm Post subject: |
|
|
Copy the output to a file, move it with usb stick to your working internet computer and paste its content, some user take screenshots also.
Anything but i wouldn't trust that log as "returns the same as", your ifconfig doesn't show ipv6 ip, while the log do see ipv6 support, your ifconfig show 19.6Gib of RX packets, pretty impressive for a card that doesn't work ; your ifconfig show irq17 that seems quiet low for a modern computer (too few irq, too many devices = conflict), while the log show irq68 that looks more sane.
And for many users udev-210 mean building a new kernel to enable FHANDLE, if you've done that, i wouldn't put that in category "without any obvious reason"
So, better gave real results if you really need help, i'm pretty sure even a seer would complain if you go on with a "read my friend hand and tell me my future".
Ok, here's the only blind help i would provide you :
setting the card ip by hand doesn't work : you must use a IP in the range of your router, you must have a dns server set to query internet, and must set its gateway too. |
|
Back to top |
|
|
jorro n00b
Joined: 09 Feb 2010 Posts: 20
|
Posted: Mon Mar 10, 2014 10:40 pm Post subject: |
|
|
Hi Krinn,
thanks for the reply...
As I said in the initial post, the IP settings are correct (or at least were), they worked and the network stopped working in a runnig system where nothing changed, except that I was working on it as a user browsing the web, programming in emacs. I set the correct IP, the DNS servers were set, the gateway was set. I could not even ping the machine next to it with its IP.
Quote: | And for many users udev-210 mean building a new kernel to enable FHANDLE, if you've done that, i wouldn't put that in category "without any obvious reason" |
I do not quite get that. I can tell you now that I went through the upgrade process using the howtos and put everything required in the kernel settings. As I wrote before, that was a few weeks ago. I cannot check the kernel config now, I am out of office till Wednesday, then I will check back and send all the logs directly copied from the not working machine...
Cheers, Jorro |
|
Back to top |
|
|
jorro n00b
Joined: 09 Feb 2010 Posts: 20
|
Posted: Wed Mar 12, 2014 2:42 pm Post subject: |
|
|
So, here we are:
ifconfig eth0 with dhcp:
Code: | eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 169.254.247.3 netmask 255.255.0.0 broadcast 169.254.255.155
inet6 <IP6_ADDRESS> prefixlen 64 scopeid 0x20<link>
ether <MAC_ADDRESS> txquelen 1000 (Ethernet)
RX packets 3189612 bytes 430711915 (410.7 MiB)
RX errors 0 dropped 3 overruns 0 frame 0
TX packets 17590 bytes 3065841 (2.9 MiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
device interrupt 17
|
ifconfig eth0 with fixed IP:
Code: | eth0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet <CORRECT_IP> netmask <CORRECT_NM> broadcast <COORECT_BC>
inet6 <IP6_ADDRESS> prefixlen 64 scopeid 0x20<link>
ether <MAC_ADDRESS> txquelen 1000 (Ethernet)
RX packets 181 bytes 32011 (31.2 KiB)
RX errors 0 dropped 3 overruns 0 frame 0
TX packets 35 bytes 6645 (6.4 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
device interrupt 17
|
When I do /etc/init.d/net.eth0 restart (no difference between dhcp and fixed IP),
dmesg -w:
Code: | [182310.052734] tg3 0000:05.0 irq 68 for MSI/MSI-X
[182310.121151] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[182313.205334] tg3 0000:05.0 eth0: Link is up at 1000 Mbps, full duplex
[182313.205337] tg3 0000:05.0 eth0: Flow control is on for TX and on for RX
[182313.205346] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
|
udev is version 210, according to http://wiki.gentoo.org/wiki/Udev/upgrade#udev_208_to_210 , I need some kernel settings, so:
grep '\(CONFIG_FHANDLE=\|CONFIG_NET=\|CONFIG_DMIID=\)' /usr/src/linux/.config
Code: | CONFIG_FHANDLE=y
CONFIG_NET=y
CONFIG_DMIID=y
|
lpci | grep Ethernet
Code: | 05:00.0 Ethernet controller: Broadcom Corporation NetXtreme BCM5761 Gigabit Ethernet PCI2 (rev 10)
|
uname -a
Code: | Linux <HOSTNAME> 3.10.25-gentoo #2 SMP PREEMPT Mon Mar 10 12:34:56 CET 2014 x86_64 Intel(R) Xeon(R) CPU W3565 @ 3.20GHz GenuineIntel GNU/Linux
|
Anyone any idea? |
|
Back to top |
|
|
creaker l33t
Joined: 14 Jul 2012 Posts: 651
|
Posted: Wed Mar 12, 2014 3:00 pm Post subject: |
|
|
jorro wrote: |
Anyone any idea? |
I had exactly the same unusual IP address, netmask and broadcast assigned by dhcp. And the same behaviour with manual IP assigning. In my case it was a cable (RG-45 connector) fault.
May be it is your case too? |
|
Back to top |
|
|
jorro n00b
Joined: 09 Feb 2010 Posts: 20
|
Posted: Wed Mar 12, 2014 3:27 pm Post subject: |
|
|
Thanks for the hint, I just tried out a new cable, no help. What confuses me is that it stopped working out of nowhere and it still works with the live USB stick of SystemrescueCD. |
|
Back to top |
|
|
jorro n00b
Joined: 09 Feb 2010 Posts: 20
|
Posted: Fri Mar 14, 2014 2:21 pm Post subject: |
|
|
So... today I had some free time and went back to that desktop. I made en eix-sync and an emerge -DavuN world while being on a chroot from a live USB stick (updated to udev-211 on the way). While that was running, I configured and build a new kernel from scratch (updateing to gentoo-sources-3.12.13). I rebootet to the system on disk and found dhcp still not working, but if I set it to a fixed IP, everything is fine. I am writing from it right now. I am not sure if this is solved... I do not trust it completely as dhcp has still some problems. I will check into that... |
|
Back to top |
|
|
jorro n00b
Joined: 09 Feb 2010 Posts: 20
|
Posted: Thu May 22, 2014 12:49 pm Post subject: |
|
|
I didn't write here for a while because the machine was working at first. After some time it stopped working again, but I did not have much time on my hand. So I quickly installed the new desktop I had here anyway and used it. I wanted to come back to fix this issue when I had more time. But now, the new desktop started to behave weird too. It has a different IP, everything new hardware. And it did the same. First time two weeks ago, I was working on it and suddenly could not get any connection to the web anymore. It was set up to use fixed IPs and no dhcp and worked like that for a month or so. I wasn't able to open any webpage, ping anything or connect to anything but I could ping the machine on the desk next to it by using the IP address. So, I updated everything and went through a gcc update ending up recompiling everything and building a new kernel using genkernel (the kernel change alone did not help anything). And after the desktop compiled over night, I rebooted and had network again.
So... that was two weeks ago on Wednesday. Yesterday (again Wednesday), almost exactly the same time, I was away from the desktop for an hour, when I came back the network was down again. dhcp does not get an IP when I configure it for the interface, and with setup IP (same setting as for the last 2 months), I do not get any connection. In the logs, the only thing I see in the time when I was gone is this:
Code: |
May 21 16:04:14 [avahi-daemon] Invalid response packet from host x.x.x.x
|
and the usual crontab entries. Also this entry does not seem to be anything special, I see many like that before. I can see the exact time I came back (the moment I plugged in my phone) and know when approximately I left, no log entries apart from this in that time.
And again, booting with a live stick and the network is perfectly running.
Has anyone any ideas what I can do? I do not want to recompile everything every two weeks... And I am out of ideas, looked around already again, found nothing. Seems to be something with my setup, but I did not reuse any configuration when installing the new desktop, the installed packages are similar, both use lvm for usr, var, tmp, and opt, both use cups, and so on. |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|