Open Stack

On May 19, 2014, at 1:04 PM, <Rob_Hirschfeld at Dell.com> <Rob_Hirschfeld at Dell.com> wrote:

> All,
>  
> I would very much appreciate hearing from the committee about positives and negatives raised during the summit about DefCore.  To me, silence means we have to slow down and get more feedback.  So even a “+1” on this email would be helpful. 
>  
> Overall, the people like the approach and results.  I got only favorable comments about how we are determining core and the way we are driving the process.  I heard no objections to the criteria or resulting small-core.

Overall a lot of support for the effort. I was pleased with the amount of discussion at the summit.

>  
> Here are the issues raised (feel free to reply with elaboration):
>  
> 1.       Swift capabilities are scored as core.  This will force the question of which parts of Swift are designated.  If 0% then other object stores just need API compatibility.  If 100% then we’re going to have to address vendors who want to ship Swift alternatives without the code.   At this point, there’s nothing in between.

I think this warrants some discussion. I had a number of talks with people about it. I think another option is to separate object store as its own “core” separate from compute/block/network. It would allow a clearer answer for those wanting to deploy only an object store solution. I think we will ultimately have to choose a “least worst” option here as there is no clear cut answer that meets everyone’s needs


>  
> 2.       Nova Docker is an interesting use case because Docker does not use Glance or Cinder in the same way as libvirt Nova does.  This means that a Docker cloud would not pass the core tests for Glance or Cinder as currently structured.

Seems to be some discussion about how Docker ultimately integrates into things in general. I’m not sure I heard a clear direction from those involved. But, it is an interesting case to consider.

>  
> 3.       We need more technical community input into capabilities grouping.  Troy did great but we need to transfer ownership.

+1

>  
> 4.       We should use the Ice House tempest even for Havana because of structural improvements made to the latest Tempest code.

+1 

>  
> Rob
> ______________________________
> Rob Hirschfeld
> Sr. Distinguished Cloud Solution Architect
> Dell | Cloud Edge, Data Center Solutions
> cell +1 512 909-7219 blog robhirschfeld.com, twitter @zehicle
> Please note, I am based in the CENTRAL (-6) time zone
>  
> _______________________________________________
> Defcore-committee mailing list
> Defcore-committee at lists.openstack.org
> http://lists.openstack.org/cgi-bin/mailman/listinfo/defcore-committee

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/defcore-committee/attachments/20140519/f156f869/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 841 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.openstack.org/pipermail/defcore-committee/attachments/20140519/f156f869/attachment-0001.pgp>