ServerManager :: HA provision failure with new params.

[Ritam Gangopadhyay]

with cluster.json in new params format HA provisioning gets stuck at pre_exec_vnc_galera_started

{
    "cluster": [
        {
            "id": "cluster5sanity",
            "email": "<email address hidden>",
            "parameters": {
                "internal_vip": "10.204.217.176",
                "provision": {
                    "contrail": {
                        "orchestrator": "openstack",
                        "kernel_upgrade": true,
                        "enable_lbass": true,
                        "ha": {
                            "haproxy_flag": false
                        },
                        "database": {
                            "database_dir": "/home/cassandra",
                            "minimum_diskGB": 32
                        },
                        "analytics": {
                            "analytics_data_ttl": 168
                        },
                        "control": {
                            "encap_priority": "MPLSoUDP,MPLSoGRE,VXLAN",
                            "router_asn": 64512,
                            "external_bgp": "[('mx2', '10.204.216.245')]"
                        },
                        "config": {
                            "use_certs": false
                        }
                    },
                    "openstack":{
                        "openstack_passwd_list": ["contrail123"],
                        "keystone_admin_user": "admin",
                        "keystone_admin_tenant": "admin",
                        "keystone": {
                            "admin_token": "contrail123",
                            "admin_password": "contrail123"
                        },
                        "neutron": {
                            "shared_secret":""
                        },
                        "multi_tenancy": true,
                        "enable_ceilometer": true,
                        "ha": {
                            "internal_vip": "10.204.217.176"
                        }
                    }
                }
            }
        }
    ]
}

[Abhay Joshi (abhayj)]

Ritam,

Is this working with old params? Have you checked?

[Abhay Joshi (abhayj)]

I see one thing incorrect in your config with new params.

Please remove :

"neutron": {
    "shared_secret":""
}

[Abhay Joshi (abhayj)]

Removing blocker tag IFF problem seen only with new params. Nitish will look into it. If old params are working, then we should not have this bug as blocking. In 3.0, we have NOT made new params mandatory.

[Ritam Gangopadhyay (ritam)]

cluster.json used for this bug:-

{
    "cluster": [
        {
            "id": "cluster5sanity",
            "email": "<email address hidden>",
            "parameters": {
                "provision": {
                    "contrail": {
                        "orchestrator": "openstack",
                        "kernel_upgrade": true,
                        "enable_lbass": true,
                        "ha": {
                            "haproxy_enable": false
                        },
                        "database": {
                            "directory": "/home/cassandra",
                            "minimum_diskGB": 32
                        },
                        "analytics": {
                            "data_ttl": 168
                        },
                        "control": {
                            "encapsulation_priority": "MPLSoUDP,MPLSoGRE,VXLAN",
                            "router_asn": 64512,
                            "external_bgp": "[('mx2', '10.204.216.245')]"
                        },
                        "config": {
                            "use_certs": false
                        }
                    },
                    "openstack":{
                        "openstack_passwd_list": ["contrail123"],
                        "keystone": {
                            "admin_user": "admin",
                            "admin_tenant": "admin",
                            "admin_token": "contrail123",
                            "admin_password": "contrail123"
                        },
                        "neutron": {
                            "shared_secret":""
                        },
                        "multi_tenancy": true,
                        "enable_ceilometer": true,
                        "ha": {
                            "internal_vip": "10.204.217.176"
                        }
                    }
                }
            }
        }
    ]
}

[Ritam Gangopadhyay (ritam)]

"openstack_passwd_list": ["contrail123"],

 in cluster json was causing the issue. This needs to be avoided.