Comment 0 for bug 1623374

SM:R3.1build33: Provisioning gets stuck with mongo connection error for database node in a MultiInterface Cluster

1) Install R3.1 build 33 Kilo Ubuntu SM
2) reimage the target with ubuntu 14.04.04. reimaged successfully
3) Add the contrail package ubuntu R3.1 build 33 kilo, Initiate Provision. Observe that the provision gets stuck at ha_proc completed, and not proceeding with database role cluster with following error

Sep 13 23:39:02 nodec33 puppet-agent[1268]: Could not retrieve catalog from remote server: Error 400 on SERVER: Could not find resource 'Notify[contrail::profile::mongodb - {primary_db_ip} = 192.168.100.3]' for relationship on 'Exec[exec_mongo_connection]' on node nodec33.englab.juniper.net

Setup details :

SM : 10.204.217.160

root@nodej5:~# server-manager status server
{
    "server": [
        {
            "id": "nodec57",
            "ip_address": "10.204.221.61",
            "mac_address": "00:25:90:C5:58:6E",
            "status": "provision_started"
        },
        {
            "id": "nodec33",
            "ip_address": "10.204.221.59",
            "mac_address": "00:25:90:C4:82:28",
            "status": "haproxy_completed"
        },
        {
            "id": "nodec35",
            "ip_address": "10.204.221.58",
            "mac_address": "00:25:90:C4:7A:70",
            "status": "haproxy_completed"
        },
        {
            "id": "nodea4",
            "ip_address": "10.204.221.60",
            "mac_address": "00:25:90:A5:3B:12",
            "status": "provision_started"
        }

root@nodej5:~# server-manager show server --select id,cluster_id,roles,ip_address
{
    "server": [
        {
            "cluster_id": "cluster_multi",
            "id": "nodec57",
            "ip_address": "10.204.221.61",
            "roles": [
                "compute"
            ]
        },
        {
            "cluster_id": "cluster_multi",
            "id": "nodec33",
            "ip_address": "10.204.221.59",
            "roles": [
                "webui",
                "config",
                "database",
                "control",
                "collector"
            ]
        },
        {
            "cluster_id": "cluster_multi",
            "id": "nodec35",
            "ip_address": "10.204.221.58",
            "roles": [
                "control",
                "config",
                "openstack"
            ]
        },
        {
            "cluster_id": "cluster_multi",
            "id": "nodea4",
            "ip_address": "10.204.221.60",
            "roles": [
                "compute"
            ]
        }
    ]
}

Notes :

1) Issue seen with only Multi Interface setup; (tested with kilo,mitaka)
2) Issue not seen in Single node, Open stack HA setups;
3) Seen in only ubuntu target. Centos Multi interface provisioning goes through fine