F9 potential service network bug?

[Dan Williams]

On Fri, 2008-05-16 at 23:47 -0400, Paul Wouters wrote:
> On Fri, 16 May 2008, Chris Ricker wrote:
> 
> > > I've personally tested static IP addressing with static DNS servers with
> > > anaconda on the F9 release LiveCD, and it worked fine for me through
> > > multiple tests.  What's not working for you?  Can you attach the ifcfg
> > > files anaconda generated right after you installed?
> >
> > Multiple people (including me) have put non-working examples in the mother
> > of all NM bugzilla tickets ;-)
> >
> > Have you tried non-LiveCD?
> 
> Mine was from a non-LiveCD. The ifcfg files generated look fine:
> 
> # Intel Corporation 82557/8/9/0/1 Ethernet Pro 100
> DEVICE=eth0
> BOOTPROTO=static
> HWADDR=00:d0:b7:85:f6:03
> ONBOOT=no
> DHCP_HOSTNAME=bofh.xelerance.com

Bill figured this one out.

> # Realtek Semiconductor Co., Ltd. RTL-8169 Gigabit Ethernet
> DEVICE=eth1
> BOOTPROTO=static
> HWADDR=00:0c:76:bc:9b:d1
> IPADDR=193.110.157.17
> NETMASK=255.255.255.248
> ONBOOT=yes

And for this one, I'm building:

http://koji.fedoraproject.org/koji/taskinfo?taskID=618281

which will pull the gateway out of /etc/sysconfig/network, which seems
to be where the non-LiveCD installer puts the gateway.  Not sure why the
livecd and the non-livecd cases are different, but this update should
fix it.

Dan

> eth0 has (or had) no link during install.
> 
> To be it looked like despite being confgured with BOOTPROTO=static, the
> avahi daemon and NetworkManager got started, and one of them started
> doing dhcp. The static IP did appear as an additiona IP on the eth1
> interface, but since the DHCP was the main address, my public ip address
> changed.
> 
> Paul
>