[clue-tech] upstart

[Dennis J Perkins]

On Mon, 2008-09-08 at 22:04 -0600, Nate Duehr wrote:
> Dennis J Perkins wrote:
> 
> > Some of LSB seems irrelevent. Standardizing certain libraries?  It might
> > be more productive to simply say that these libraries must be present in
> > an LSB-compliant system.  Many distros are aleady using those many of
> > those libraries already.
> 
> That is THE MOST important part to any ISV trying to create BINARY 
> software that works on any distro!  Linux/Unix is generally hostile to 
> binary software in general, so it's not too much of a surprise to see a 
> talented Unix/Linux admin as yourself surprised that it's important.

Sorry, I'm not an admin.  I program PLC's. :)

Most distros seem to be using the same library packages, which is why
trying to standardize them seems a waste of time.  It's not like trying
to standardize many versions of C.  The only library differences might
be the versions in use.

> I agree with you however, that it shouldn't demand the libraries are 
> used, just available and working...
> 
> > The justification is that there is little or no difference between a
> > device event and a time event.  I don't know if I agree with that.  That
> > depends on how it would be implemented.  It might mean that a
> > crontab-style program provides the user interface, and upstart is the
> > back end.  It might not take much more code to do this.  Maybe.
> 
> It takes a LOT of retraining for sysadmins, though... relearning 
> something other than cron after almost 20 years of using it, and the new 
> thing adds nothing a good scripter (sysadmin) couldn't do with cron and 
> sysV?  ONLY if it makes the hard things EASY for more than just 
> sysadmins... maybe then it's worth it.  If it's just to change things 
> for change sake, it's not.

Upstart's job files might be simpler than regular bootscripts.  That's
why I might install it and play with it to see how good and easy it
actually is.  Or how many problems it has.

I haven't seen any info on how they plan to try cron-style stuff.  They
aren't even focusing on that yet.  But if crontab can talk to upstart,
maybe the interface will be the same.

> > Sometimes its a refactoring.  Instead of reinventing the wheel, refactor
> > by placing commonly used procedures into libraries.  This is one
> > advantage of desktop systems, even if some garbage gets dragged in too.
> 
> Time to dump KDE/GNOME and "refactor" the desktop.  LOL!

Think about it.  They have a common printing library, or it's
incorporated into a bigger library, so every program can provide the
same interface instead of reinventing the wheel.  The same is true for
fonts, window layout, widgets...  Now why did I even mention
desktops? :)

> > The bootscripts don't monitor the daemons they start and restart them if
> > they crash.  If cron, postfix, postgresql, etc., crash, are they
> > restarted automatically?  Maybe some of them, but probably not all of
> > them.  And if upstart's init crashes, the kernel restarts it.  Hmm, I
> > wonder what effect that would have?  No matter.  Sysvinit would have the
> > same problem.
> 
> init itself does a great job of restarting things that need to run all 
> the time, actually... /etc/inittab ... and an app written well enough 
> not to run multiple copies of itself...

If it's in inittab.  But what if it's started by a bootscript
in /etc/init.d?

> >>> There are probably other things that could be mentioned, but I think this is enough for now.
> >> So far, they're all not very compelling reasons to swap out SysV and 
> >> switch to this thing, really.
> > 
> > Maybe not yet.  And maybe initng, runit or something else will be
> > preferred by most people or distros.
> 
> If "not yet" is the answer, it shouldn't be the default then.  Leave 
> SysV init and let the devs rip it out and replace it on their boxes 
> until they have a COMPELLING reason to use the new one done for the rest 
> of us.

Ubuntu and Fedora are still running in SysV compatabiity mode, so the
traditional /etc/init.d and /etc/rc.d still works.  Inittab itself is
mostly gone.  Only the default runlevel line is left.  There are a few
tty jobs in /etc/event.d to handle the old tty lines.  At the
moment, /etc/event.d seems to only be used to replace what inittab did.
Unless you run Frugalware, which decided to push the envelope and see
how far it could go.