Wed UTC: 12:00 / China 20:00 / Eastern: 08:00 / Pacific: 05:00
https://zoom.us/j/2868829092
Participants
Chuyi Guo Byung-Woo Jun Fernando Oliveira Alessandro Gerardo D'Alessandro Andy MayerXu Yang
Agenda
- NSD and AID SD mapping
- IFA 015 UML work
Minutes
- Agenda Bashing
- IFA 015 UML work hasn't been discussed, we will wait Michela to discuss together.
- NSD and AID SD mapping
- Fred and Byung showed an analysis of ETSI model and AID model mapping, the presented analysis can be found: ETSI Package Management (SDC Enhancements)- Guilin.
- Current analysis for NSD includes attribute mapping between SOL001 and SDC org.openecomp.resource.vfc.NSD, based on the material shown by Thinh before.
In the material, there has a suggestion to use tosa.nodes.nfv.NS directly in SDC. - The modeling work for NSD can be 1)align NSD IM to v2.7.1, 2) translate it into ONAP internal data model.
- The basic issue for NSD is no initial node type for NSD in SDC. VFC is using org.openecomp.resource.vfc.NSD, but the goal is to provide a general model mapping that can be used by all other run-time components, like DCAE, which highly depend on AID model. In R7, there will have contribution to support SOL 007 in SDC.
- PNFD has works for mapping with AID model by Ericsson in SDC, but no works for VNFD. VNFD suffers challenges. In AID model, there has VF module, while ETSI doesn't, VF module and ETSI SOL 001 is not 1 to 1 mapping. The ONAP components may need VF module, but vendors don't need. The mapping work is also expected to support CNF as well.
- VLs now are vendor specific, and is differnt from ETSI in current implementation. Suggested solution is to use tosca.nodes.nfv NsVirtualLink.
- Suggest to have a talk with VFC to seek for their perspective, Byung will summary this, and send email to VFC team. There has Orchestration Scenarios Meeting on Monday to talk about ETSI alignment issues, Meeting Information can be found: Orchestration Scenarios