[Mageia-dev] ANN: For the brave. systemd v185 in cauldron updates_testing

[Colin Guthrie]

'Twas brillig, and Olav Vitters at 28/06/12 09:54 did gyre and gimble:
> On Thu, Jun 28, 2012 at 09:21:10AM +0100, Colin Guthrie wrote:
>> 'Twas brillig, and Olav Vitters at 27/06/12 22:21 did gyre and gimble:
>>> On Wed, Jun 27, 2012 at 09:31:44PM +0200, Olav Vitters wrote:
>>>> Never really tried making a plot quickly after booting, but 8min should
>>>> be enough to have every service started.
>>>
>>> Due to mysqld/mariadb not starting
>>>
>>> Loads of repeating messages in the logs:
>>> Jun 27 21:56:03 bkor systemd[1]: mysqld.service holdoff time over,
>>> scheduling restart.
>>> Jun 27 21:56:03 bkor systemd[1]: Job pending for unit, delaying
>>> automatic restart.
>>
>> Hmm, interesting. MySQL starts fine here.
> 
> Table corruption :P
> 
> | InnoDB: The log sequence number in ibdata files does not match
> | InnoDB: the log sequence number in the ib_logfiles!
> | 120627 21:22:06  InnoDB: Database was not shut down normally!
> | InnoDB: Starting crash recovery.
> | InnoDB: Reading tablespace information from the .ibd files...
> | InnoDB: Restoring possible half-written data pages from the doublewrite
> | InnoDB: buffer...
> | 120627 21:22:06  InnoDB: Assertion failure in thread 140513125668672 in
> | file fsp0fsp.c line 2874
> | InnoDB: Failing assertion: space != 0
> 
> A bit strange as I always reboot cleanly, but in any case not related to
> systemd. Could be that I never noticed this failing (haven't used the
> databases in many months).

Yeah, as none of the other services will really depend on mysql (or note
that you actively use at least) and thus systemd has probably been
diligently trying hard to start it for a while. It will eventually give
up and your systemd-analyze plot will show a nice loooooooooong bar for
this particular service :)

>>> I uninstalled mariadb and it went through. Before that tried a chkconfig
>>> mysqld off, but that didn't seem to help.
>>
>> MySQL has a native systemd unit so you should have done "systemctl
>> disable mysqld.service". Once a native unit exists the enabled/disabled
>> status of the sysvinit script is irrelevant.
> 
> chkconfig forwards things right? I'm still using chkconfig + service.

Nope. The brige is the other way around. i.e. systemctl disable will
forward TO chkconfig if there is no native systemd unit, but chkconfig
will NOT forward to systemctl if there is a native systemd unit.

Rule: Always use systemctl, forget chkconfig.

>> As a side note, how was your keyboard layout and do you have a
>> /etc/X11/xorg.conf.d/00-keyboard.conf file?
> 
> Don't have /etc/X11/xorg.conf.d/00-keyboard.conf. My keyboard layout
> is&was standard (US). I have capslock as compose key in xorg, but didn't
> check if that is still working.

Ahh if you have a US layout then that's likely not an issue you'd get
tripped up by.

>> Just FYI, having spoken to Lennart, the plan is to write a patch for
>> systemd-localed that will allow us to force a sync of locale settings
>> (and thus write the above Xorg snippet) to ensure everyone gets the
>> right keymapping. We will trigger this functionality on upgrade. This is
>> currently the only thing holding me back from pushing this. I'll try and
>> find time to write such a patch at the weekend.
> 
> Nice! Because then I can finish the GNOME upgrade :-)
> (most pending stuff is either waiting on systemd, or new
> evolution-data-server.. and cannot upgrade evolution-data-server without
> upgrading the stuff which wants a newer systemd)


Yup, I appreciate I'm holding things up here :s I'll try and get this
done ASAP.

Cheers!

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/