Comment 43 for bug 1066410

After a surprising 5 successful restarts with 'sleep 6' added into /etc/init/lighdm.conf file, I thought that some recent updates might have fixed it. I've kept an eye on the updates that were being applied and they don't look to be related but hey, the proof was in the pudding - five successful restarts.

I then commented out the 'sleep 6' and Lightdm is back to being a recalcitrant and not starting. Adding 'sleep 6' back in got it going again.

I suspect that after my hybrid HDD starts caching again that it will break due to some race condition.

For me previously, it sometime wont even start when I execute 'sudo lightdm' from the command-line. When it fails after trying to manually start it, I have a series of about 20 "Error writing to daemon: Broken pipe" messages appear on the console.

I've attached my /var/log/lightdm/x-0-greeter.log file. It looks as if the broken pipe message may have some relationship to DBUS.