[lxc-devel] lxd: Initial design specifications

[Dwight Engen]

On Fri, 7 Nov 2014 10:16:06 -0500
Stéphane Graber <stgraber at ubuntu.com> wrote:

> On Fri, Nov 07, 2014 at 09:52:34AM -0500, Dwight Engen wrote:
> > On Fri, 7 Nov 2014 00:39:19 +0000
> > Serge Hallyn <serge.hallyn at ubuntu.com> wrote:
> > 
> > > Quoting Stéphane Graber (stgraber at ubuntu.com):
> > > > On Thu, Nov 06, 2014 at 04:03:18PM -0500, Dwight Engen wrote:
> > > > 
> > > > Re-reading the examples, except for that specific one, there's
> > > > nothing that copy can do which start doesn't do too.
> > > 
> > > Yeah, Tycho and I noticed that today right after pushing code
> > > which implemented... create :)
> > > 
> > > > So I see two options there and I'm not completely sure which I
> > > > prefer to be honnest:
> > > > 1) We drop copy and we add a --later option to the start
> > > > command. 2) We drop copy and we introduce a create command
> > > > which is identical to start except that it doesn't start the
> > > > container and doesn't accept any runtime-specific argument.
> > > > 
> > > > My problem with 1) is the obvious conceptual issue that start in
> > > > that mode won't actually start anything.
> > > > 
> > > > My problem with 2) is the addition of a new command just for
> > > > what may well end up being a corner case.
> > > 
> > > So for lxc 1.0 Stéphane worked hard to remove a slew of redundant
> > > lxc-* commands.  So there's a strong desire to not err that way
> > > again.  However I think 'lxc start --later' is working too hard to
> > > avoid having 'lxc create'.  I don't think that create without
> > > start is a corner case.
> > > 
> > > I do agree the semantic overlap is unfortunate.  But having to do
> > > 'lxc-start --later' to create feels akin to going to the start
> > > menu to shut down.
> > 
> > I agree start --later is weird. We could just have create, and it
> > could have a --start option, but that doesn't seem very
> > discover-able. If we're following a traditional hypervisor model,
> > it seems they all have separate create vs. start so having start
> > actually create something is a bit different.
> 
> Ok, good. That's what Serge proposed as a pull request last night and
> which got merged, so we now have separate create and start and no more
> copy.
> 
> Create and start are basically the same syntax and should be sharing
> most of their code.
> 
> > > So I'd vote for adding create.  I'll propose a patch in a few
> > > minutes to add that to the docs.
> > > 
> > > thanks,
> > > -serge
> 
> Can I still get you to send a pull request for profile apply?

Sure thing, just have to figure out github vs git-send-email :)
 
> Thanks
>