Description of problem:
After upgrading from Fedora 14 to Fedora 15 the system boots into maintenance mode after:
Cannot add dependency job for unit autofs.target, ignoring: Unit autofs.target failed to load: No such file or directory. See system logs and 'systemctl status' for details

After various:
Unit systemd-logger.service entered failed state
systemd-logger.service start request repeated to quickly, refusing to start.

udev: converting old udev database
Ext3-fs (dm-0): using internal journal
---- [ cut here ] ----
Warning: at drivers/char/tty_io.c:1325 tty_open+0x213/0x3b7()
Hardware name: 
Modules linked in:
Pid: 137, comm: plymouthd Not tainted 2.6.36.1-7.rc1.fc15.i686 #1

the whole thing crashes. And boots into maintenance mode. This on vmware, kvm *and* on real hardware. Taking a plain vanilla kernel not created, patched and build by fedora makes the system boot on vmware, kvm *and* real hardware.

Version-Release number of selected component (if applicable):
2.6.36.1-7.rc1.fc15.i686

for later kernels:
worker did not accept message -1 (Connection refused), kill it

multiple times. After these the system boots into maintenance mode.
It is the same for all newer kernels:
2.6.38.6-26.rc1.fc15.i686
2.6.38.5-24.fc15.i686
2.6.38.2-9.fc15.i686

How reproducible:
always. Just upgrade from Fedora 14.


Steps to Reproduce:
1. Install Fedora 14
2. Update to the latest package-versions
3. Upgrade to Fedora 15
  
Actual results:
Unusable system.

Expected results:
Fedora 15 running seamlessly

Additional info:
Ferora runs if kernels are replaced with plain vanila ones from kernel.org, compiled using Fedora 14.