[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Bug#650400: netcfg: Fails to obtain an address via DHCP



On Tue 29 Nov 2011 at 15:02:30 -0400, Joey Hess wrote:

> Brian Potkin wrote:
> Nov 29 14:21:59 kernel: [   28.066457] 8139too: 8139too Fast Ethernet driver 0.9.28
> Nov 29 14:21:59 kernel: [   28.066528] 8139too 0000:00:0a.0: PCI INT A -> GSI 17 (level, low) -> IRQ 17
> Nov 29 14:21:59 kernel: [   28.067314] 8139too 0000:00:0a.0: eth0: RealTek RTL8139 at 0xe000, 00:a0:d2:a5:ad:25, IRQ 17
> Nov 29 14:21:59 kernel: [   28.067865] 8139cp: 8139cp: 10/100 PCI Ethernet driver v1.3 (Mar 22, 2004)
> ...
> Nov 29 14:22:01 check-missing-firmware: no missing firmware in /dev/.udev/firmware-missing /run/udev/firmware-missing
> Nov 29 14:22:01 kernel: [   29.842892] 8139too 0000:00:0a.0: eth0: link up, 100Mbps, full-duplex, lpa 0x45E1
> ...
> Nov 29 14:22:35 main-menu[292]: (process:4523): Sending discover...
> Nov 29 14:22:35 main-menu[292]: (process:4523): udhcpc: sendto: Network is down
> Nov 29 14:22:35 main-menu[292]: (process:4523): Read error: Network is down, reopening socket
> Nov 29 14:22:35 main-menu[292]: (process:4523): Read error: Network is down, reopening socket
> ...
> Nov 29 14:24:13 netcfg[4524]: DEBUG: dhcp_client_sigchld() called
> Nov 29 14:24:13 kernel: [  161.621068] 8139too 0000:00:0a.0: eth0: link up, 100Mbps, full-duplex, lpa 0x45E1
> ...
> Nov 29 14:24:13 netcfg[4524]: INFO: executing: ip addr add 192.168.7.90/24 broadcast 192.168.7.255 dev eth0
> ...
> Nov 29 14:24:32 main-menu[292]: (process:4523): udhcpc: sendto: Network is down
> Nov 29 14:24:32 main-menu[292]: (process:4523): Read error: Network is down, reopening socket
> 
> This looks like it starts with the ethernet link up, and goes down
> before udhcpc is started, then after several minutes it goes up again,
> and even then udhcpc fails -- perhaps the link has gone down a second
> time?

The second 'link up' is when the interface was configured statically.

> This could be anything from a faulty driver to a hardware problem to a
> cable problem.

I'm faced with a businesscard image from 19th November and a Squeeze
netinst image which configure the network in 10s with four different
network cards and drivers. On the same machines this daily image (and
the ones for the past few days) gives up. I'll look at the situation
over the next few days without ruling out a difficiency in my setup.


Reply to: