openstack_user_config duplicate veth names

[Lori Dabbs]

The inventory generation probably should detect and generate an error when there are duplicate veth names defined within openstack_user_config file.
Causes ssh errors when attempting to run playbooks lxc-container-create and the deployment node is attempting to ssh into containers - failed bez of duplicate reference to eno2 as mgmt and storage br.

root@infra1:/opt/openstack-ansible/playbooks# vi /etc/openstack_deploy/openstack_user_config.yml
---
cidr_networks:
  container: 10.150.241.0/24
  tunnel: 10.150.242.0/24
  storage: 10.150.243.0/24

used_ips:
  - "10.150.241.2,10.150.241.50"
  - "10.150.242.2,10.150.242.50"
  - "10.150.243.2,10.150.243.50"

global_overrides:
  internal_lb_vip_address: 10.150.241.10
  # The external IP is quoted simply to ensure that the .aio file can be used as input
  # dynamic inventory testing.
  external_lb_vip_address: 10.150.240.10
  tunnel_bridge: "br-vxlan"
  management_bridge: "br-mgmt"
  provider_networks:
    - network:
        container_bridge: "br-mgmt"
        container_type: "veth"
        container_interface: "eno2"
        ip_from_q: "container"
        type: "raw"
        group_binds:
          - all_containers
          - hosts
        is_container_address: true
        is_ssh_address: true
    - network:
        container_bridge: "br-vxlan"
        container_type: "veth"
        container_interface: "eno10"
        ip_from_q: "tunnel"
        type: "vxlan"
        range: "1:1000"
        net_name: "vxlan"
        group_binds:
          - neutron_linuxbridge_agent
    - network:
        container_bridge: "br-vlan"
        container_type: "veth"
        container_interface: "eno12"
        host_bind_override: "eno12"
        type: "flat"
        net_name: "flat"
        group_binds:
          - neutron_linuxbridge_agent
    - network:
        container_bridge: "br-vlan"
        container_type: "veth"
        container_interface: "eno11"
        type: "vlan"
        range: "1:1"
        net_name: "vlan"
        group_binds:
          - neutron_linuxbridge_agent
    - network:
        container_bridge: "br-storage"
        container_type: "veth"
        container_interface: "eno2"
        ip_from_q: "storage"
        type: "raw"
        group_binds:
          - glance_api
          - cinder_api
          - cinder_volume
          - nova_compute
          - swift_proxy
root@infra1:/opt/openstack-ansible/playbooks# vi /etc/openstack_deploy/openstack_user_config.yml
---
cidr_networks:
  container: 10.150.241.0/24
  tunnel: 10.150.242.0/24
  storage: 10.150.243.0/24

used_ips:
  - "10.150.241.2,10.150.241.50"
  - "10.150.242.2,10.150.242.50"
  - "10.150.243.2,10.150.243.50"

global_overrides:
  internal_lb_vip_address: 10.150.241.10
  # The external IP is quoted simply to ensure that the .aio file can be used as input
  # dynamic inventory testing.
  external_lb_vip_address: 10.150.240.10
  tunnel_bridge: "br-vxlan"
  management_bridge: "br-mgmt"
  provider_networks:
    - network:
        container_bridge: "br-mgmt"
        container_type: "veth"
        container_interface: "eno2"
        ip_from_q: "container"
        type: "raw"
        group_binds:
          - all_containers
          - hosts
        is_container_address: true
        is_ssh_address: true
    - network:
        container_bridge: "br-vxlan"
        container_type: "veth"
        container_interface: "eno10"
        ip_from_q: "tunnel"
        type: "vxlan"
        range: "1:1000"
        net_name: "vxlan"
        group_binds:
          - neutron_linuxbridge_agent
    - network:
        container_bridge: "br-vlan"
        container_type: "veth"
        container_interface: "eno12"
        host_bind_override: "eno12"
        type: "flat"
"/etc/openstack_deploy/openstack_user_config.yml" 143L, 3091C 2,1 Top
  infra1:
    ip: 10.150.240.6
  infra2:
    ip: 10.150.240.7
  infra3:
    ip: 10.150.240.8

identity_hosts:
  infra1:
    ip: 10.150.240.6
  infra2:
    ip: 10.150.240.7
  infra3:
    ip: 10.150.240.8

compute_hosts:
  compute1:
    ip: 10.150.240.9

storage_hosts:
  storage1:
    ip: 10.150.240.5
# container_vars:
# cinder_backends:
# limit_container_types: cinder_volume
# lvm:
# volume_group: cinder-volumes
# volume_driver: cinder.volume.drivers.lvm.LVMVolumeDriver
# volume_backend_name: LVM_iSCSI

log_hosts:
  haproxy_log:
    ip: 10.150.240.10

network_hosts:
  infra1:
    ip: 10.150.240.6
  infra2:
    ip: 10.150.240.7
  infra3:
    ip: 10.150.240.8

haproxy_hosts:
  haproxy_log:
    ip: 10.150.240.10

[Jean-Philippe Evrard (jean-philippe-evrard)]

This is a wishlist item for feature addition.

On another topic, are you sure you want these defined that way?

container_interface: "eno12"
host_bind_override: "eno12"

[Lori Dabbs (ldabbs)]

Thank you for pointing this out.
We were using a guideline of a previous configuration done on a different environment. We are trying to build our own private cloud with minimal knowledge.
I will make note on the virtual network interfaces within the configuration file and make sure they all have unique NIC IDs.

Lori Dabbs
UT Health San Antonio
IMS / EPI Sr System Admin
210.567.2102

-----Original Message-----
From: <email address hidden> [mailto:<email address hidden>] On Behalf Of Jean-Philippe Evrard
Sent: Monday, November 21, 2016 7:34 AM
To: Dabbs, Lori <email address hidden>
Subject: [Bug 1643056] Re: openstack_user_config duplicate veth names

This is a wishlist item for feature addition.

On another topic, are you sure you want these defined that way?

container_interface: "eno12"
host_bind_override: "eno12"

--
You received this bug notification because you are subscribed to the bug report.
https://bugs.launchpad.net/bugs/1643056

Title:
  openstack_user_config duplicate veth names

Status in openstack-ansible:
  New

Bug description:
  The inventory generation probably should detect and generate an error when there are duplicate veth names defined within openstack_user_config file.
  Causes ssh errors when attempting to run playbooks lxc-container-create and the deployment node is attempting to ssh into containers - failed bez of duplicate reference to eno2 as mgmt and storage br.

  root@infra1:/opt/openstack-ansible/playbooks# vi /etc/openstack_deploy/openstack_user_config.yml
  ---
  cidr_networks:
    container: 10.150.241.0/24
    tunnel: 10.150.242.0/24
    storage: 10.150.243.0/24

  used_ips:
    - "10.150.241.2,10.150.241.50"
    - "10.150.242.2,10.150.242.50"
    - "10.150.243.2,10.150.243.50"

  global_overrides:
    internal_lb_vip_address: 10.150.241.10
    # The external IP is quoted simply to ensure that the .aio file can be used as input
    # dynamic inventory testing.
    external_lb_vip_address: 10.150.240.10
    tunnel_bridge: "br-vxlan"
    management_bridge: "br-mgmt"
    provider_networks:
      - network:
          container_bridge: "br-mgmt"
          container_type: "veth"
          container_interface: "eno2"
          ip_from_q: "container"
          type: "raw"
          group_binds:
            - all_containers
            - hosts
          is_container_address: true
          is_ssh_address: true
      - network:
          container_bridge: "br-vxlan"
          container_type: "veth"
          container_interface: "eno10"
          ip_from_q: "tunnel"
          type: "vxlan"
          range: "1:1000"
          net_name: "vxlan"
          group_binds:
            - neutron_linuxbridge_agent
      - network:
          container_bridge: "br-vlan"
          container_type: "veth"
          container_interface: "eno12"
          host_bind_override: "eno12"
          type: "flat"
          net_name: "flat"
          group_binds:
            - neutron_linuxbridge_agent
      - network:
          container_bridge: "br-vlan"
          container_type: "veth"
          container_interface: "eno11"
          type: "vlan"
...