Re: rc improvements (wanted?)

From: Peter Ross <Peter.Ross_at_alumni.tu-berlin.de>
Date: Sun, 20 Jul 2008 02:32:09 +1000 (EST)
[Resending - maybe I used the wrong mail address which is not subscribed 
to -current? Does it matter? Anyway, the message did not make it to the 
list, it seems, and I did not get a notice.. strange]
---------- Forwarded message ----------
Date: Sun, 20 Jul 2008 01:07:32 +1000 (EST)

On Sat, 19 Jul 2008, Robert Noland wrote:

> On Fri, 2008-07-18 at 22:19 -0700, Doug Barton wrote:
> > Bernd Walter wrote:
> > > Speaking about small systems, where startup time is more a problem than
> > > on 08/15 desktop and server systems.
> > > What I would love to see is that scripts like moused, ypserv, lpt, etc
> > > are not started if the services are disabled.
> > 
> > That wold be a neat trick, how do you propose we accomplish it? (no, 
> > I'm not being snide.)
> > .. 
> > One way you could do this is to have /etc/rc.d/active and 
> > /etc/rc.d/inactive (and probably an /etc/rc.d/system for critical 
> > stuff that most people shouldn't touch). Then you could have a 
> > vipw-like system to allow users to edit rc.conf that would move the 
> > scripts to the right directory. Of course, this would be fraught with 
> > potential for problems. :)
> 
> I almost hate to toss this out there, but what about a sys v type rc?

Usually the scripts are running run_rc_command.

This does a checkyesno for the rcvar variable which is usually 
${name}_enable. In most of the cases it uses set_rcvar to achieve this.

If so, /etc/rc could evaluate ${name}_enable (it already knows them via 
/etc/rc.conf) before actually calling a script.

There are some irregular names amongst the name of the rcvar variable.

I am not 100% sure whether same of the scripts set ${name}_enable vaiables 
implicitly. That could complicate things. I could not find evidence by 
browsing through them.

I also do not know whether there are scripts that actually do something 
valuable before calling run_rc_command.

At the moment it looks to me that these excemptions could be dealt with by 
adapting the scripts so they meet the standard (running only when 
${name}_enable is set).

I only looked through some of the /etc/rc.d scripts en detail (+ some 
greps for rcvar etc. in the directory) so it needs some more 
investigation.

If it works it avoids messing around with symlinks or moving scripts 
around, and it reduces the scripts that actually run.

As a sys admin I really like the BSD way of having everything relevant to 
my system in one /etc/rc.conf. It is very convenient.

Regards
Peter
Received on Sat Jul 19 2008 - 17:40:03 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:33 UTC