kube-manager not creating virtual-router reference for kube-dns virtual-machine object on resetting contrail config

[Vedamurthy Joshi]

R4.0 Build 3041

On clearing cassandra using the reset_config flags in the config daemons( so that objects can be populated from fresh), it is seen that for the kube-dns virtual-machine object, no virtual-router reference is created. So this object is not seen in agent at all

Ex :
{
"virtual-machine": {
"fq_name": [
"kube-dns-1024139455-bg2wg"
],
"uuid": "fb0d71de-fb2e-11e6-93a3-525400010001",
"href": "http://10.204.217.194:8082/virtual-machine/fb0d71de-fb2e-11e6-93a3-525400010001",
"virtual_machine_interface_back_refs": [
{
"to": [
"default-domain",
"kube-system",
"pod-kube-dns-1024139455-bg2wg-79e3610c-fb42-11e6-92a9-525400010001"
],
"href": "http://10.204.217.194:8082/virtual-machine-interface/79e3610c-fb42-11e6-92a9-525400010001",
"attr": null,
"uuid": "79e3610c-fb42-11e6-92a9-525400010001"
}
],
"perms2": {
"owner": "cloud-admin",
"owner_access": 7,
"global_access": 0,
"share": []
},
"id_perms": {
"enable": true,
"description": null,
"creator": null,
"created": "2017-02-25T10:09:21.219265",
"user_visible": true,
"last_modified": "2017-02-25T10:09:21.241178",
"permissions": {
"owner": "cloud-admin",
"owner_access": 7,
"other_access": 7,
"group": "cloud-admin-group",
"group_access": 7
},
"uuid": {
"uuid_mslong": 18090240480727929000,
"uuid_lslong": 10638437265510040000
}
},
"display_name": "kube-dns-1024139455-bg2wg",
"annotations": {
"key_value_pair": [
{
"value": "K8S:POD",
"key": "device_owner"
},
{
"value": "{\"name\": \"kube-dns\", \"pod-template-hash\": \"1024139455\", \"component\": \"kube-dns\", \"kubernetes.io/cluster-service\": \"true\", \"tier\": \"node\", \"k8s-app\": \"kube-dns\"}",
"key": "labels"
}
]
},
"name": "kube-dns-1024139455-bg2wg"
}
}

[Vedamurthy Joshi (vedujoshi)]

It was because vrouter objects themselves were removed.