Bug: OSM does not find mgmt nertwork

[mpeuster]

Bug

Upon instantiation of the SFB demo service (which worked with an early version of OSMr4 + vim-emu) OSM reports the following error:

ERROR Waiting for management IP address reported by the VIM. Updating VNFRs: ns_update_vnfr: Not fou
nd member_vnf_index=2 vdur=vdu01 interface=v2-out at RO info

However, all three VNFs containers are started and running.

LCM Module reports

ERROR lcm.ns ns.py:1060 Task ns=ff7d10c4-0347-43e5-9245-4e27962dd112 instantiate=620d42ad-d951-403b-b814-74c782bb4b59 Exit Exception while 'Waiting for management IP address reported by the VIM. Updating VNFRs': ns_update_vnfr: Not found member_vnf_index=2 vdur=vdu01 interface=v2-out at RO info

Reproduce

• Everything is based on this repository
• Used VNFDs and NSDs: osm-demo-service-cdn
• Follow this walkthrough.

Helper:

• vim-emu logs: docker logs vim-emu
• RO logs 1: docker logs osm_ro.1.ulj81w4qnqcyaol100iqz4dnz
• RO logs 2: docker exec osm_ro.1.ulj81w4qnqcyaol100iqz4dnz cat /var/log/osm/openmano.log
• LCM docker logs osm_lcm.1.ze4898swowexp5ah2goqt2924

[mpeuster]

Solution:

Problem was that vnf0 had an input interface defined and vnf2 had an output interface defined but both interfaces were NOT used in the NSD!

So this error seems to always point to unused connection points defined in VNFDs but not used in the NSD!