Question about the 'K' and 'S' in script names in /etc/rc.d/rc{0, 6}.d

Bruce Dubbs bruce.dubbs at
Tue Jul 19 12:59:22 PDT 2011

DJ Lucas wrote:
> On 07/19/2011 12:20 AM, Bruce Dubbs wrote:
>> DJ Lucas wrote:

>> Actually, I haven't done a manual change of run level is several years.
>>    I like to start in RL3 and stay there until shutdown.  It's easy to
>> type startx and the boot is a lot faster.  For those who are command
>> line impaired, I guess a change from RL5 to RL3 is sometimes needed, but
>> C-A-D for reboot and shutdown for halt is all I ever use.
>> I can't remember the last time I used RL1 or RL2 for any reason.

> Traitor! That's the perfect argument for upstart (systemd too?). ;-)

I don't agree.  I think the main argument for those is to boot faster 
via parallel bootscript execution.   To me, this is a solution looking 
for a problem.  I instrumented the startup for my LFS system:

Total time according to the kernel timer
to the login prompt                     : 13.61 seconds
Total time from 1st to last bootscript  : 9 seconds.
Time for udev to start                  : 5 seconds
Time for checkfs (6 partitions)         : 2 seconds

How much time is systemd going to save?  I looked for benchmarks, but 
couldn't find any.  I did find a lot of arguments saying parallel would 
be faster, but none that stated how much.

The really big time consumer for most systems is bringing up X and a 
heavy duty (e.g. Gnome, KDE) graphical environment.  The tasks there are 
  tied to one startup script via xdm or equivalent.  If speedup is 
wanted, concentrate on the process that takes the time.

This is in the same vein as using dash instead of bash because it's 
faster.  How much?  Not enough to notice without instrumentation.

upstart/systemd trade complexity for negligible gains.  I don't want to 
see them in LFS.


   -- Bruce

More information about the lfs-dev mailing list