I think it's worth repeating that there has been quite a bit of development on
imagefactory since the 1.0 release currently used by conductor.
With imagefactory 2.0, the "magic provider name" issue is gone. We have also
changed the way we provide status, providing both a simplified status string meant for the
UI as well as the full log detail when digging deeper is necessary. While we endeavor to
provide more meaningful status messages in 2.0, with the number of moving parts and
subcomponents we use, I cannot promise every situation has been corrected.
Along with the changes mentioned above,
- we made much needed changes to the REST interface to make it more sane
- we improved the CLI to make it more useable
- we introduced a plugin API and converted our builder code to plugins
- we added a plugin for OpenStack
The only problem is that, with all of the change we introduced (especially in the REST
interface), conductor needed changes to work with the post 1.x imagefactory. We have been
working with the developers on the conductor team that are leading that effort and it
sounds as though it will not be too much longer before this integration is completed and
the above changes can be used.
-steve
On Oct 23, 2012, at 4:47 AM, Martin Povolny <mpovolny(a)redhat.com> wrote:
>> Log in, adding a provider
>>
>>
http://192.168.122.174:3000/providers
>> #1 error: "Provider name must start with 'ec2-'"
>>
>> WTF? why? the UI allows any name! (was told, this is fixed somewhere)
>
> I'd be curious to know if this was actually fixed, and if the fix was
> that we were no longer required to have "magic" names that indicated
> provider type.
As I understand it, yes, ImageFactory should be fixed. But I'm not sure
about it, will try to find out.