[systemd-bugs] [Bug 54247] systemd boot messages in the login prompt

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=54247

Zbigniew Jedrzejewski-Szmek <zbyszek at in.waw.pl> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |NEW

--- Comment #17 from Zbigniew Jedrzejewski-Szmek <zbyszek at in.waw.pl> ---
Hm, actually my previous comment was totally wrong. We simply delay the idle
jobs for a maximum of IDLE_TIMEOUT_USEC (which is defined to 5 s). It is
relatively easy to have services which take more than 5 s to start, and then
gettys will be launched while other things are still being started.

We could bump those 5 s to something bigger, but there are huge downsides - if
something goes wrong, logging in will be impossible for longer, potentially
indefinitely. So setting the timeout to something huge or infinity seems risky.

Type=idle is a hack, and it shows here. It's seems unlikely that we can find an
IDLE_TIMEOUT_USEC value which will satisfy everybody. Actually 5 s seems to be
a reasonable value.

A more complicated option would be to add an additional target, let's say
multi-user-post.target, which would be After=multi-user.target,
Requires=multi-user.target, and make getty services be part of it. It would be
a solution with the similar downsides as simply increasing the
IDLE_TIMEOUT_VALUE, unlikely to be worth it.

Dunno, we could also export IDLE_TIMEOUT_USEC as a config option. But it is a
very low-level thingy, and not something that should be a part of the external
interface. Maybe IDLE_TIMEOUT_USEC should instead be scaled by the number of
jobs and processor speed or having a SSD/non-SSD disk. Again, probably not
worth the hassle for a cosmetic issue.

I'm resetting this to NEW, maybe somebody can come up with a solution.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/systemd-bugs/attachments/20130624/44bb6663/attachment.html>