[Mageia-dev] Broken database problem solved but not all well

[Colin Guthrie]

'Twas brillig, and Anne Wilson at 17/04/12 07:24 did gyre and gimble:
> On 16/04/12 16:58, Colin Guthrie wrote:
>> OK, so this is incorrect, (prefdm *is* started, but perhaps some other
>> jobs had to be cancelled instead.
>>
>> Can you do this:
>>
>>
>> 1. Add the following the to kernel command line when booting normally:
>> "systemd.log_level=debug systemd.log_target=kmsg"
>>
>> 2. When you boot and it gets stuck, simply go to tty2 copy
>> /var/log/dmesg to some file and then somehow give it to me :)
>>
> Hope this helps :-)

OK so there are definitely ordering cycles here:

[   18.312187] systemd[1]: Activating default unit: default.target
[   18.312219] systemd[1]: Trying to enqueue job
graphical.target/start/replace
[   18.312874] systemd[1]: Found ordering cycle on basic.target/start
[   18.312891] systemd[1]: Walked on cycle path to sockets.target/start
[   18.312905] systemd[1]: Walked on cycle path to syslog.socket/start
[   18.312917] systemd[1]: Walked on cycle path to basic.target/start
[   18.312930] systemd[1]: Breaking ordering cycle by deleting job
syslog.socket/start

And herein lies the problem probably. This might affect a few other
things that require to be after syslog.socket. (dbus itself is after
syslog.target so this could be affected here which would explain why
kdm failed to start properly as it couldn't connect to dbus)


[   18.313131] systemd[1]: Looking at job plymouth-quit.service/stop
conflicted_by=yes
[   18.313146] systemd[1]: Looking at job plymouth-quit.service/start
conflicted_by=no
[   18.313160] systemd[1]: Fixing conflicting jobs by deleting job
plymouth-quit.service/start
[   18.313186] systemd[1]: Looking at job prefdm.service/start
conflicted_by=no
[   18.313199] systemd[1]: Looking at job prefdm.service/stop
conflicted_by=no
[   18.313212] systemd[1]: Fixing conflicting jobs by deleting job
prefdm.service/stop
[   18.313237] systemd[1]: Looking at job getty at tty1.service/stop
conflicted_by=yes
[   18.313250] systemd[1]: Looking at job getty at tty1.service/start
conflicted_by=no
[   18.313263] systemd[1]: Fixing conflicting jobs by deleting job
getty at tty1.service/start
[   18.313400] systemd[1]: Installed new job graphical.target/start as 1


These are normal conflicting jobs and this bit is expected behaviour.


All of this should be fixed when we update all the legacy initscripts.

In the mean time if you go into /etc/init.d/ and do:

grep -c "INIT INFO" *| grep -vE
"(halt|killall|mandrake_.*|reboot|single|functions):" | grep :0$


This should show you what initscripts you have on your system that do
not have LSB header information.

If you let me know the results of that command I'll look specifically
at this one so I can push a fix for you sooner rather than latter
(tho' all will get fixed before RC).

Col

-- 

Colin Guthrie
colin(at)mageia.org
http://colin.guthr.ie/

Day Job:
  Tribalogic Limited http://www.tribalogic.net/
Open Source:
  Mageia Contributor http://www.mageia.org/
  PulseAudio Hacker http://www.pulseaudio.org/
  Trac Hacker http://trac.edgewall.org/