On Tue, 2012-01-24 at 14:53 +0100, Michal Schmidt wrote:
On 01/23/2012 09:39 PM, Adam Williamson wrote:
> systemd 37 didn't have a problem booting. I suppose the
> large number of errors slows udev start down so much that it hits some
> kind of timeout in 38 that isn't in 37, maybe?
v38 introduced the journal and removed the old systemd-kmsg-syslogd
bridge. As a side effect of the change, the ordering of syslog.socket to
the rest of the system is not quite correct now. If you're unlucky, you
will see journald blocking for extended periods of time until rsyslog
starts. And other processes (in your case: udev) will block on journald.
I noticed the SELinux relabeling process was affected by the same problem.
We have to fix that in systemd.
That sounds plausible - kay even tried putting the same bad udev rules
onto his systems and couldn't reproduce the bug, so it does look like a
case of having to be very (un)lucky with timing.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | identi.ca: adamwfedora
http://www.happyassassin.net