Description of problem: Anaconda always fails to bring up eth0 via dhcp. Version-Release number of selected component (if applicable): Anaconda shipped on Fedora 10 32bit. How reproducible: Every Time Steps to Reproduce: 1. Boot Fedora 10 32bit install media and append ks="http://example.com/kickstart" (Or use koan/cobbler) 2. Wait for the network portion to start. Actual results: Anaconda fails to bring up eth0 via dhcp. Expected results: Dhcp successfully completes. Additional info: Manually hitting enter ("Retry") will instantly allow Anaconda to continue successfully. Apparently not card dependent, bug on at least the following cards. 02:08.0 Ethernet controller: Intel Corporation 82801DB PRO/100 VE (LOM) Ethernet Controller (rev 81) 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168B PCI Express Gigabit Ethernet controller (rev 01) It's a blocker bug for me, as I am trying to kickstart a lab. Having to hit enter on each machine does not allow unattended installs.
Are you able to try rawhide and see if this is still happening? I ask because this bug sounds familiar to a problem we solved recently in rawhide. https://fedoraproject.org/wiki/Releases/Rawhide
I can respin a revisor image to try. What would the appropriate way to include just Rawhide anaconda into a respin?
Including just anaconda from rawhide won't work because a lot has changed in anaconda and the things it requires are different. It would probably be easiest to just try boot.iso from rawhide. Those would normally appear in these directories: http://download.fedora.redhat.com/pub/fedora/linux/development/i386/os/images/ http://download.fedora.redhat.com/pub/fedora/linux/development/x86_64/os/images/ http://download.fedora.redhat.com/pub/fedora/linux/development/ppc/os/images/ If one is missing, it means the tree build for that architecture did not complete for the night (for one reason or another).
I'm still unable to reproduce this on rawhide over the past week.