Interjecting in the middle of the discussion which is always dangerous but … If you do a recursive grep in the oom repo of “image:” and maybe only include those with onap in the image name you will ge

#helm #docker

What we do in Azure on AKS is provide openstack parameters for a target Devstack instance that is emulating an openstack based remote data center but if you are only doing k8 and CNFs you wont need th

Probably can leave them as defaults since those robot functions won’t work anyway if there isn’t a heat engine to use for VNF virtual machine deployments. Brian

SO is populating a null for that value in the call to SDNC. You need to look at SO and see why it isn’t seeing the right mapping from the model to that value in the variable. The GET to aai is returni

#cnf

Take a look at the robot log.html that is attached to the following jira. Scroll down to see how robot creates the policy updates after the VM is instantiated. Then you can look at the files that are

#controlloop

https://docs.openstack.org/keystone/ussuri/api_curl_examples.html check and make sure that a) with project scope payload that you get the endpoints section and b) in multicloud you are setting the pro

#msb #multicloud

authenticate fails: did you check that the url, login and password , (and if v3 which it must be ) project domain and user domain are correct ? Brian

#msb #multicloud

The multi-cloud team may have a better answer but the Guilin version of multicloud.pike includes additional variables like ssl_enabled and msbprotocol to set up https on port 443 in addition to the ms

#msb #multicloud

I see weirdness all over the logs but the one that might be the critical one is the Multicloud getting an HTTP 500 back from 10.108.10.41 - - [08/Jan/2021:14:11:47 +0000] "POST /CloudOwner/k8sregionfo

Praveen, You can see an example of setting the path to the caFile in the DANOS VES collector (among others). In that example VNF we pass the caFile full path as a parameter to the vpp-measurement-repo

#ves

It looks like you have a Basic Auth header in your curl. I think even though you specified a -u that header is taking a priority. remove the basic auth and use -u or create a basic auth header with th

#ves

Jim, I have not setup the ingress controller so cant be much help. The readthedocs page has some guidance that you would need to customize for Azure with both the ingress controller and the dns setup

I think you have type and name flipped. vnf-type should be the model name and vnf-name should be an instancename. Although I could be interpreting your variable names incorrectly. Brian "vnf-name": "{

#guilin

As you can see in the log – SO cant talk to your cloud. Its getting an HTTP 404 back from the HTTP connection to openstack. Looks like you are using the keystone v2 config instead of the keystone v3 s

Check the SO openstack adapter for errors. The error message seems to indicate the source of the error is OPENSTACK. "extSystemErrorSource" : "OPENSTACK", but I’m not totally familiar with this error

Remembered that I had some postman queries for cluster state. In my environment peer voting status is healthy with all peer’s reporting true. sdnc ssl nodeport is 30267 https://{{sdnc_port}}/jolokia/r

Correcting some misleading typo’s Brian

If you look inside the guilin-sdnc-2 you will like see opendaylight errors in the karaf.log. I would delete the pods that have shared errors in karaf.log and make sure healthcheck for sdnc passes. the

Sounds like one of your sdnc containers is down (perhaps dev-sdnc-2). kubectl -n onap get pod | grep sdnc Sdnc failing healthcheck means you won’t be able to deploy a VNF anyway. I’d restart the faile

Looks like GLOBAL_INJECTED_PRIVATE_KEY is the variable in /share/config/robot_properties.py that contains the path to the keyfile and in your case it is null. Populate it in the robot container config