Description of problem: the halt script calls out to kexec prior to calling. This seemingly causes access to the root NFS to timeout. When the script runs in debug mode, I get the timeout when this command is run. Line 220 in /etc/init.d/halt [ -f /poweroff -o ! -f /halt ] followed eventually by NFS timeouts. The remote client is still pingable. The network isn't gone. If I comment out the line right above, [ -n "$kexec_command" ] && $kexec_command -e >& /dev/null the reboot works perfectly. Version-Release number of selected component (if applicable): system-config-netboot-0.1.41-1.FC6.noarch How reproducible: always
Tracked this down to the specifics on the kexec call. Initscripts version initscripts-8.45.7-1 The call to the kexec command at the end of the halt /etc/init.d/halt script should be changed from [ -n "$kexec_command" ] && $kexec_command -e >& /dev/null to [ -n "$kexec_command" ] && $kexec_command -e -x >& /dev/null This leaves up the network interfaces which is important on a NFS mounted root.
This seems to be a dupe of #223932 on Enterprise Linux. Any chance at a backport of it to FC6 as errata? It's for a netboot FC6 system which is ill behaved on F7 still.
Maybe, will have to see what else is pending.