A similar problem I had was the result of the 30 seconds a switch port listens for spanning tree traffic before enabling when the link comes up. This was enough time to interfere with UDP communication (dhcp and rpc). Enabling spantree portfast on the switch port fixed all of the boot and install problems we were having related to DHCP and RPC timeouts. PXE boot agents, pump, dhclient and ethtool will all cause the link to reset, triggering the 30 second listen on spanning tree supporting switches. -----Original Message----- From: Raphaël 'SurcouF' Bordet [mailto:surcouf@xxxxxxxxxxxx] Sent: Monday, October 25, 2004 7:39 AM To: Kickstart-list@xxxxxxxxxx Subject: Re: [SOLVED] Network Kickstart FC2 installation on PXE and NFS Raphaël 'SurcouF' Bordet wrote: [...] > I follow your thread: I've do same configuration that yours except I > don't have this options set but nfs server isn't called... > > /tftpboot/linux-install/pxelinux.cfg/default: [...] > label redhat > KERNEL /rhel3/vmlinuz > APPEND initrd=/rhel3/initrd.img ksdevice=eth0 network \ > ks=nfs:192.168.0.50:/kickstart/as1bdbi1-ks.cfg > __END__ I've resolved by using http instead of nfs: it's working fine. But I don't understand why nfs doesn't work: I can mount installation directory share where's RedHat tree but anaconda can't mount my kickstart share (and I can mount this by hand)... label redhat KERNEL /rhel3/vmlinuz APPEND initrd=/rhel3/initrd.img ksdevice=eth0 network \ ks=http://192.168.0.50/kickstart/ -- Raphaël 'SurcouF' Bordet
