On 01/10/13 00:07, Jaromir Coufal wrote: > Just BTW: > > I know that lot of folks were watching the youtube stream > (http://youtu.be/m3y6uD8yKVQ), so please feel free to give any feedback > you have to this thread. I believe that this is good way to proceed > forward and how to make things flexible enough to support various types > of deployments. > > Looking forward to any feedback
my 2c: * I like LGroup. You can still create an LGroup where all nodes in a physical rack are in the same LGroup (i.e. what we have now in Rack). However, it also means you can divide a given physical rack of servers into two or more LGroups (eg TOR switch setup to give you two rack-local subnets). Gives much better capacity planning. For Tuskar... currently a 'Rack' is just a list of node_ids ... so the concept should actually map fine - I think? * hardware profile on a resource class... nice idea... i like the auto matching of profiles to particular nodes though obviously thats a while away yet. In tuskar we'd need to expand the resource class definition and operations for capturing the hardware profiles. Right now only racks have this notion of 'aggregate capacity' and resource class doesn't expose the total aggregate 'total resource capacity' or aggregate 'total instance capacity'. marios > > Thanks > -- Jarda > > On 2013/30/09 13:57, Jaromir Coufal wrote: >> Hi everyone, >> >> based on Tuskar's merger with TripleO and upstream feedback on Tuskar, >> when I was thinking about processes and workflows there, I got into >> some changes which I think that are important for us, because they >> will help us to achieve better flexibility and still having ability >> for easy scaling. >> >> I wanted to do just walkthrough the wireframes but I think that it >> will raise up some discussion around Classes and Racks, so my thought >> was to merge both together (wireframes + concepts discussion). >> >> At this meeting I'd like to get you familiar with my thoughts and get >> into some wireframes which will explain the ideas more. I hope that we >> will get into discussion around changes (not just UI but API as well). >> >> The scope which we will be talking about is Icehouse. >> >> I'll be posting link into #tripleo IRC channel. >> I'd like to record the whole session, so if anybody cannot attend, it >> should be available for you later. >> >> (Please note that Google Hangout has limited number of 10 >> participants, so if you consider just watching, please use youtube >> stream - link will be posted here when available.) >> >> Thanks >> -- Jarda >> >> >> _______________________________________________ >> OpenStack-dev mailing list >> [email protected] >> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev > > > > > _______________________________________________ > OpenStack-dev mailing list > [email protected] > http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev > _______________________________________________ OpenStack-dev mailing list [email protected] http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
