Here is the latest version of the diagram. Please let me know of any corrections or changes. Jerry
Hi Jerry, G-lambda-K will support both of NSIv2 and NSIv1 in KDDILabs and JGNX domain. Regards, Takahiro On 2012/11/01 18:10, Jerry Sobieski wrote:
Here is the latest version of the diagram. Please let me know of any corrections or changes. Jerry
-- Takahiro Miyamoto <tk-miyamoto@kddi.com> KDDI Corporation TEL: +81-80-5945-9004
Hi, Any OpenNSA instance can run as an aggregator, so also the ones at the UvA and Starlight. But this works only for v1, so I'm not sure whether it is useful to add this capability to the diagram at all... Jeroen. On 1 Nov 2012, at 10:10, Jerry Sobieski <jerry@nordu.net> wrote:
Here is the latest version of the diagram. Please let me know of any corrections or changes. Jerry <Supercomputing2012-SLC AutomatedGOLE v2 Diag.pdf>
Here is V3 diagram. Please note that unless your GOLE/Network has an "A", you are considered a uPA. Only aggregators will be expected to do multi-domain segmentation. And AIST's G-LAMBDA-A is still the only bi-lingual aggregator. The bi-colored GOLEs are "bi-lingual"- they have both a v1 WS endpoint and a v2 WSendpoint announced in the respective topologies. It is assumed that these versions will be coordinating the resource allocations for the covered GOLE... i.e. that the v1 NSA knows if/when the v2 NSA allocates a VLAN or bw so that they don't walk on each other. PLease let me know if you will be running both versions - and if they do/do not coordinate resources. And of course any other corrections are invited. Jerry On 11/1/12 5:10 AM, Jerry Sobieski wrote:
Here is the latest version of the diagram. Please let me know of any corrections or changes. Jerry
Hello, It is ok when it comes to pionier/geant. Additional details to be entered in the topology: 1) pionier.net (the one that is attached to netherlight): provider endpoint - http://194.132.53.174:9280/nsi/ConnectionProvider 2) pionier.junos.exp.net (the one that is connected to pionier.net) provider endpoint - http://194.132.53.174:9282/nsi/ConnectionProvider 3) geant.net provider endpoint - http://gnscdemo.geant.net:8080/nsi/ConnectionProvider br michal -----Oryginalna wiadomość----- From: Jerry Sobieski Sent: Thursday, November 01, 2012 10:59 AM To: nsi-wg@ogf.org ; automatedgole-pilot@internet2.edu Subject: Re: [Nsi-wg] AutoGOLE diagram for SC. Here is V3 diagram. Please note that unless your GOLE/Network has an "A", you are considered a uPA. Only aggregators will be expected to do multi-domain segmentation. And AIST's G-LAMBDA-A is still the only bi-lingual aggregator. The bi-colored GOLEs are "bi-lingual"- they have both a v1 WS endpoint and a v2 WSendpoint announced in the respective topologies. It is assumed that these versions will be coordinating the resource allocations for the covered GOLE... i.e. that the v1 NSA knows if/when the v2 NSA allocates a VLAN or bw so that they don't walk on each other. PLease let me know if you will be running both versions - and if they do/do not coordinate resources. And of course any other corrections are invited. Jerry On 11/1/12 5:10 AM, Jerry Sobieski wrote:
Here is the latest version of the diagram. Please let me know of any corrections or changes. Jerry
_______________________________________________ nsi-wg mailing list nsi-wg@ogf.org https://www.ogf.org/mailman/listinfo/nsi-wg
participants (4)
-
Jeroen van der Ham -
Jerry Sobieski -
michalb -
Takahiro Miyamoto