> > So, (most of) the milestones actually do relate to
> runlevels, but the milesto
> > ne names need to be read carefully.
>
> You've got it right. The names "single-user",
> "multi-user", and
> "multi-user-server" were chosen because that's how
> init(1M) has
> described the S, 2, and 3 runlevels.
OK, thanks. Nice to finally get it straight. ;-)
> Perhaps that's the best way to explain...
> Each explicit milestone (single-user, multi-user,
> multi-user-server)
> runs the services which in Solaris 9 and below were
> run in
> rcS, rc2, and rc3, respectively. In Solaris 10,
> SMF introduces
> two new milestones: none, and all. 'none' reflects
> a system with
> no services started. 'all' reflects a system with
> all enabled
> services started, including those that were not
> included in one
> of the explicit milestones.
> oes that help clarify?
Immensely. So, "all" is the default, which actually equates to something like
"runlevel 3+" - a level that didn't really exist before.
> Nope, sorry. "all" is the default milestone. This
> is specified
> in svc.startd(1M), where it can be overridden.
Gotcha. Sorry, but I'm at work, so I wasn't able to look at this on my Sol10
boxes before posting, and therefore kept assuming incorrectly.
> Should be. There's a bug filed that it isn't. (I
> can track the # down
> later if you want.)
I think that might be Christine's.
> I know I helped someone write this table recently,
> but can't remember
> where it got published.
>
> Milestone Run Level
> none N/A
> milestone/single-user S
> milestone/multi-user 2
> milestone/multi-user-server 3
> all 3
The table's great. You may want to add a "+" on the last 3 ("all") to keep it
clear from multi-user-server's relation to runlevel 3. Just a suggestion from a
de-confused guy. :-)
> Thanks for helping us work through this to make it
> actually useful and
> understandable.
Heck, thank YOU (and David) for helping to clarify this for me. I think
Christine is right in her comment above, that there's probably quite a number
of people that have this all turned around. I've been focusing on other areas,
so I took some of the inaccurate (or otherwise unclear) material on SMF at face
value.
I've put the link to your blog entry in the Big Doc List (under the Admin area,
if I remember correctly). If you ever post a new version of the entry, or some
other doc that helps clarify it like you did here, I'd be more than happy to
update the link. Just let me know. :-)
Rainer
This message posted from opensolaris.org
