Hi Freek, No, it is the later "... a subset of a topology from a given domain that consists of the resources and relations that have been changed/added/removed since a given timestamp?" The idea was that we needed to verify topology because we may pull it from a third party server. To verify topology, we proposed "signing" the topology in full. The issue is that if you update part of the topology (e.g. add a new port), do you just submit the partial change and sign it, in which case, when you redistribute it, you'd have to keep the original topology file with the original signature distinct from the signed partial topology update. To simplify things, we said that when there is a change in the topology, the author must regenerate the whole topology again and sign it in it's entirety. In essence, a complete (domain) topology file is the atomic unit of information for the topology service. Hope that helps. - Chin On Thu, Jan 23, 2014 at 11:27 AM, Freek Dijkstra <Freek.Dijkstra@surfsara.nl
wrote:
In a off-list discussion, the "topology updates (i.e. partial topology documents)" mentioned in the presentation of Chin at OGF40 came up. (https://redmine.ogf.org/dmsf_files/13191)
I hope Chin of Inder can clarify the following:
What does such "partial topology" entail?
* Is that the collection of full topologies from those domains whose topology has updated since a given timestamp?
or
* Is that a subset of a topology from a given domain that consists of the resources and relations that have been changed/added/removed since a given timestamp?
During the presentation I assumed the later, later someone (I forgot who) convinced me it was the former. The former seems a lot more simple to implement, but re-reading the slides after Miroslav's questions I'm in doubt again.
Freek
-- Chin Guok NOC: (510) 486-7600 Network Engineer (800) 333-7638 ESnet Network Engineering Group (AS293) Lawrence Berkeley National Laboratory