Open Stack

I think the heat's version of nested containers is slightly different.
Maybe it would be good to sync up in one of the meetings to align the
stuff. The original Donabe idea is pre-heat and we presented this to TOSCA
in early 2012 and they liked it a lot. So I wouldnt be surprised if all
these things look similar at a 30K feet level :)

Regarding the Curvature backend: it is based on pure Openstack APIs and
authenticated via keystone. So one could take Curvature and just point it
to *any* pure openstack implementation and it should work.


On Thu, May 9, 2013 at 2:45 AM, Steven Hardy <shardy at redhat.com> wrote:

> On Wed, May 08, 2013 at 10:31:52PM +0000, Gabriel Hurley wrote:
> > Here's the roadmap on this topic as it stands:
> >
> > D3.js is in the process of being incorporated, and the team at Cisco
> that built Curvature is interested in combining some of the client-side
> know-how they've developed by building Curvature with the existing editable
> network topology code the folks at NTT and NEC built in Horizon, and
> combining all of that with what Heat wants to do to produce something
> AWESOME and very much in the Curvature style but built on the existing
> OpenStack APIs rather than the Ruby backend (Donabe).
> >
> > Fortunately the "nested stacks"/"recursive containers" work is separate
> from the interface work, and that's primarily the part that was written in
> Ruby (Donabe). Heat is investigating that area with other DSLs such as
> TOSCA/OASIS (forgive me if I'm wrong there), and we'll eventually all
> arrive in the same place.
>
> As previously stated[1] Heat already supports nested/recursive stacks, and
> I'm still not clear if "recursive containers" is the same thing or not, my
> assumption currently is that it is.
>
> The DSL/Providers discussion is about allowing users to specify custom
> resource types via templates (which is really just an extension to our
> current nested stack interface), so maybe that does map more closely to the
> definition of "recursive containers", or does it really mean inheritance?
>
> As you say Gabriel, hopefully over time the terminology and concepts will
> converge.
>
> [1]
> http://lists.openstack.org/pipermail/openstack-dev/2013-April/007327.html
>
> _______________________________________________
> OpenStack-dev mailing list
> OpenStack-dev at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
>



-- 
-Debo~
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20130509/543d2401/attachment.html>