My mind was tricked into saying level instead of layer ins some places - I fix them below. Sorry. John On Aug 24, 2009, at 4:15 PM, John Vollbrecht wrote:
I have some comments on the schema - I talked a bit with Jeroen about these since he is in AA this month. I like the schema and that it is defining things so they can be discussed. I particularly like the definition of adaptation as being between ports at different layers - this explains it for me and implies that adaptation takes place between ports at different layers of the same device.
I have other comments -
1) Jeroen says that there will be an introduction, and that it will define network- I just mention this. 2) I think that level / layer/ should be a top level element. I think the reason it isn't there is that there was some confusion about whether the initial schema would be single or multiple level. /ayer/ 3) Subnet - the concept from G.800, 805 and 8080 should be included. Perhaps as a top level element or as a special group - I am not sure. 4) Ports should have relationship with groups, especially with subnetworks. They also have relationship with layer. 5) I think that there could be virtual ports similar to virtual nodes, so an implemented by relation would be appropriate. Jeroen may write this up more, but 8080 show virtual ports (SNPPs) and IDCs implement multiple next hops using VLAN groups on some edge ports.
some smaller points - 6) I think link capacity should not be the only attribute in the schema. Perhaps something like "A Link MAY have additional attributes describing the performance, ownership, policy". A separate sentence might say that attributes that describe throughput MUST be in bytes per second rather than bits per second.
7) Link definition seems circular when describing it as a link :)
John
On Aug 21, 2009, at 3:55 PM, Jeroen van der Ham wrote:
Hello,
I've just submitted a new revision of the schema document to the subversion repository. Here's the changelog:
- Added a short description of the Location and Lifetime objects - Added a section enumerating all Relations. I've not called them "Relations" throughout the text. If need be, we can always make the names of Relations implicit in the syntax. - Added a section on Service, describing the way we currently see adaptations? I welcome comments! - The group section probably needs more work.
I've updated the UML schema diagram to the latest version. I've also added a note saying that this is not intended as endorsement or finalizing, merely trying to reflect the current status.
- Changed the Link group to Bidirectional Link as per the diagram.
The latest version of the PDF can be found in the subversion repository. Unfortunately the web access to it seems broken at the moment, so I'm attaching it to this mail.
Jeroen. <nml-base.pdf>_______________________________________________ nml-wg mailing list nml-wg@ogf.org http://www.ogf.org/mailman/listinfo/nml-wg