yahoo-eng-team team mailing list archive

[Assaf Muller <amuller@xxxxxxxxxx>]

This is a documentation issue. Since the Icehouse days I believe,
Neutron allows you to configure external connectivity in one of two
ways:

1) Configure the external bridge name in l3_agent.ini. This means that
'qg' external router devices are plugged directly in to the bridge, the
OVS agent doesn't manage the bridge nor does it set up flows in case the
segmentation type for the external network is 'vlan'. This means that
you must set up a Linux VLAN device over the bridge with the external
network's VLAN ID.

2) Set the external bridge name in l3_agent.ini to ''. Configure bridge
mappings in the OVS agent so that it maps the external network's
physical_network to a bridge (Typically still called br-ex). Then the
OVS agent manages the bridge and sets up flows for VLAN translation.

** Changed in: neutron
       Status: New => Invalid

** Changed in: neutron
     Assignee: ugvddm (271025598-9) => (unassigned)

-- 
You received this bug notification because you are a member of Yahoo!
Engineering Team, which is subscribed to neutron.
https://bugs.launchpad.net/bugs/1530038

Title:
  when external_network_bridge is set to  br-ex ovs does not create flow
  to EXT access

Status in neutron:
  Invalid

Bug description:
  Under  /etc/neutron/l3_agent.ini  when configuring external_network_bridge = br-ex 
  OVS does not create dump-flow for this vlan so we get external access. 

  version : # rpm -qa |grep neutron 
  openstack-neutron-7.0.1-2.el7ost.noarch
  python-neutronclient-3.1.0-1.el7ost.noarch
  python-neutron-7.0.1-2.el7ost.noarch
  openstack-neutron-openvswitch-7.0.1-2.el7ost.noarch
  openstack-neutron-common-7.0.1-2.el7ost.noarch
  openstack-neutron-ml2-7.0.1-2.el7ost.noarch
  [root@puma06 ~(keystone_admin)]# rpm -qa |grep openvswitch 
  python-openvswitch-2.4.0-1.el7.noarch
  openstack-neutron-openvswitch-7.0.1-2.el7ost.noarch
  openvswitch-2.4.0-1.el7.x86_64

  Installed by packstack OSP-8

  step to reproduce : 
  1. Deploy with packstack OSP-8 , set ETX bridge parameter -->  CONFIG_NEUTRON_L3_EXT_BRIDGE= br-ex 
  2. setup ENV : 
  # neutron net-create external_network --provider:network_type=vlan  --provider:segmentation_id=181 --provider:physical_network physnet --router:external
  # neutron subnet-create external_network 10.35.166.0/24 --disable-dhcp --gateway 10.35.166.254  --allocation-pool start=10.35.166.1,end=10.35.166.100 
  # neutron net-create int_net 
  # neutron subnet-create int_net 192.168.1.0/24 --dns_nameservers list=true 10.35.28.28 --name int_sub
  # neutron router-create Router_eNet
  # neutron router-interface-add Router_eNet subnet=int_sub
  # neutron router-gateway-set Router_eNet external_network

  3. ovs-ofctl dump-flows br-ex
  NXST_FLOW reply (xid=0x4):
   cookie=0x0, duration=57707.344s, table=0, n_packets=21, n_bytes=1638, idle_age=60, priority=2,in_port=2 actions=drop
   cookie=0x0, duration=57707.394s, table=0, n_packets=1233857, n_bytes=85172234, idle_age=0, priority=0 actions=NORMAL

  try to ping EXT network (8.8.8.8)  -- no connectivity
  4. When we change external_network_bridge = provider and creating the ENV again (new router & new routr ports )
  I get EXT access and the flow created : 
  # ovs-ofctl dump-flows br-ex
  NXST_FLOW reply (xid=0x4):
   cookie=0x0, duration=105.955s, table=0, n_packets=5, n_bytes=402, idle_age=97, priority=4,in_port=2,dl_vlan=2 actions=mod_vlan_vid:181,NORMAL
   cookie=0x0, duration=58222.490s, table=0, n_packets=39, n_bytes=3174, idle_age=105, priority=2,in_port=2 actions=drop
   cookie=0x0, duration=58222.540s, table=0, n_packets=1244856, n_bytes=85931346, idle_age=0, priority=0 actions=NORMAL

To manage notifications about this bug go to:
https://bugs.launchpad.net/neutron/+bug/1530038/+subscriptions