[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [Rdo-list] port qg-xxxxxx of external gateway goes to br-int instead of br-eth2







On 6 באוג׳ 2015, at 20:30, Dan Sneddon <dsneddon redhat com> wrote:

On 08/06/2015 11:21 AM, Dan Sneddon wrote:
On 08/06/2015 03:34 AM, ICHIBA Sara wrote:
Hello guys,

I'm trying to add a second external network to my openstack. I followed
the documentation, created a new router and set the new gateway. I
can't ping my external gateway from my router namespace  and I can ping
this gateway from my server.

I kept comparing the configuration of my new external gateway (which
doesn't work in neutron) and my oldest one (which keeps working) and
noticed that when I set a gateway to my router this creates a port with
a name that bigins with qg. Normally this port has to be assigned to
the external bridge (in my case it should be the new external bridge
br-eth2). BUT what I saw is that this port qg-xxx is mapped instead to
the br-int.

For more explanation please see my confs and some command outputs:


[root OScontroller templates(keystone_admin)]# neutron net-external-list
+--------------------------------------+-------------------+------------------------------------------------------+
| id                                   | name              |
subnets                                              |
+--------------------------------------+-------------------+------------------------------------------------------+
| 59c6f54f-26e3-4360-8a05-1a63285c846c | public            |
6744a5b3-1a33-42ae-8ae7-1467ebdc1b13 192.168.5.0/24
<http://192.168.5.0/24>  |
| 7ee9f199-2c22-42a9-a6af-1b06c2d62a35 | ruckus_management |
cba357ee-f9a7-49f3-89e4-9f5d864747cd 192.168.31.0/24
<http://192.168.31.0/24> |
+--------------------------------------+-------------------+------------------------------------------------------+







****this router with its gateway are working***

[root OScontroller templates(keystone_admin)]# ip netns exec
qrouter-e8243f85-4c56-47bd-a1ee-40724a861dc6 ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
   link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
   inet 127.0.0.1/8 <http://127.0.0.1/8> scope host lo
      valid_lft forever preferred_lft forever
   inet6 ::1/128 scope host
      valid_lft forever preferred_lft forever
25: qr-2b96e8b6-38: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc
noqueue state UNKNOWN
   link/ether fa:16:3e:b5:54:4d brd ff:ff:ff:ff:ff:ff
   inet 10.0.0.1/24 <http://10.0.0.1/24> brd 10.0.0.255 scope global
qr-2b96e8b6-38
      valid_lft forever preferred_lft forever
   inet6 fe80::f816:3eff:feb5:544d/64 scope link
      valid_lft forever preferred_lft forever
26: qr-9e50bd2e-fa: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc
noqueue state UNKNOWN
   link/ether fa:16:3e:fc:06:9b brd ff:ff:ff:ff:ff:ff
   inet 20.0.0.1/24 <http://20.0.0.1/24> brd 20.0.0.255 scope global
qr-9e50bd2e-fa
      valid_lft forever preferred_lft forever
   inet6 fe80::f816:3eff:fefc:69b/64 scope link
      valid_lft forever preferred_lft forever
27: qg-cd45a565-0b: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc
noqueue state UNKNOWN
   link/ether fa:16:3e:20:81:60 brd ff:ff:ff:ff:ff:ff
   inet 192.168.5.100/24 <http://192.168.5.100/24> brd 192.168.5.255
scope global qg-cd45a565-0b
      valid_lft forever preferred_lft forever
   inet 192.168.5.101/32 <http://192.168.5.101/32> brd 192.168.5.101
scope global qg-cd45a565-0b
      valid_lft forever preferred_lft forever
   inet 192.168.5.103/32 <http://192.168.5.103/32> brd 192.168.5.103
scope global qg-cd45a565-0b
      valid_lft forever preferred_lft forever
   inet6 fe80::f816:3eff:fe20:8160/64 scope link
      valid_lft forever preferred_lft forever






************* this is the new router and gateway that I configured and
they are not working
[root OScontroller templates(keystone_admin)]# ip netns exec
qrouter-b3c51abb-4a14-4af2-ae72-b41e7cba4e84 ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
   link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
   inet 127.0.0.1/8 <http://127.0.0.1/8> scope host lo
      valid_lft forever preferred_lft forever
   inet6 ::1/128 scope host
      valid_lft forever preferred_lft forever
38: qg-6406f3e3-72: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc
noqueue state UNKNOWN
   link/ether fa:16:3e:63:48:64 brd ff:ff:ff:ff:ff:ff
   inet 192.168.31.70/24 <http://192.168.31.70/24> brd 192.168.31.255
scope global qg-6406f3e3-72
      valid_lft forever preferred_lft forever
   inet6 fe80::f816:3eff:fe63:4864/64 scope link
      valid_lft forever preferred_lft forever




[root OScontroller templates(keystone_admin)]# ovs-vsctl show
9b3e4cc7-6761-483e-a538-1a132734d1a5
   Bridge br-tun
       Port br-tun
           Interface br-tun
               type: internal
       Port "vxlan-c0a80523"
           Interface "vxlan-c0a80523"
               type: vxlan
               options: {df_default="true", in_key=flow,
local_ip="192.168.5.34", out_key=flow, remote_ip="192.168.5.35"}
       Port patch-int
           Interface patch-int
               type: patch
               options: {peer=patch-tun}
   Bridge "br-eth2"
       Port "phy-br-eth2"
           Interface "phy-br-eth2"
               type: patch
               options: {peer="int-br-eth2"}
       Port "eth2"
           Interface "eth2"
       Port "br-eth2"
           Interface "br-eth2"
               type: internal
   Bridge br-int
       fail_mode: secure
       Port "qr-9e50bd2e-fa"
           tag: 1
           Interface "qr-9e50bd2e-fa"
               type: internal
       Port "tap61f76964-85"
           tag: 1
           Interface "tap61f76964-85"
               type: internal
       Port "qvo0952a802-99"
           tag: 2
           Interface "qvo0952a802-99"
       Port "qvo5cc221dc-e8"
           tag: 1
           Interface "qvo5cc221dc-e8"
       Port "tap599b2f77-21"
           tag: 4095
           Interface "tap599b2f77-21"
               type: internal
       Port "tapa383b7ae-99"
           tag: 2
           Interface "tapa383b7ae-99"
               type: internal
       Port "tapbea1a763-0b"
           tag: 2
           Interface "tapbea1a763-0b"
               type: internal
       Port "qr-2b96e8b6-38"
           tag: 2
           Interface "qr-2b96e8b6-38"
               type: internal
       Port int-br-ex
           Interface int-br-ex
               type: patch
               options: {peer=phy-br-ex}
       Port "qg-6406f3e3-72"
           tag: 9
           Interface "qg-6406f3e3-72"
               type: internal
       Port patch-tun
           Interface patch-tun
               type: patch
               options: {peer=patch-int}
       Port "int-br-eth2"
           Interface "int-br-eth2"
               type: patch
               options: {peer="phy-br-eth2"}
       Port br-int
           Interface br-int
               type: internal
   Bridge br-ex
       Port "qg-cd45a565-0b"
           Interface "qg-cd45a565-0b"
               type: internal
       Port br-ex
           Interface br-ex
               type: internal
       Port "eth0"
           Interface "eth0"
       Port phy-br-ex
           Interface phy-br-ex
               type: patch
               options: {peer=int-br-ex}
   ovs_version: "2.3.1"
[root OScontroller templates(keystone_admin)]#



************************fichiers de conf*********************

[root OScontroller neutron]#  cat l3_agent.ini | grep -v ^# | grep -v ^$
[DEFAULT]
debug = False
interface_driver =neutron.agent.linux.
interface.OVSInterfaceDriver
use_namespaces = True
gateway_external_network_id =
handle_internal_only_routers = True
external_network_bridge =
metadata_port = 9697
send_arp_for_ha = 3
periodic_interval = 40
periodic_fuzzy_delay = 5
enable_metadata_proxy = True
router_delete_namespaces = False
agent_mode = legacy
allow_automatic_l3agent_failover=False



root OScontroller neutron]# cat plugin.ini | grep -v ^# | grep -v ^$
[ml2]
type_drivers = vxlan
tenant_network_types = vxlan
mechanism_drivers =openvswitch
[ml2_type_flat]
[ml2_type_vlan]
[ml2_type_gre]
[ml2_type_vxlan]
vni_ranges =10:100
vxlan_group =224.0.0.1
[securitygroup]
enable_security_group = True


[root OScontroller openvswitch]# cat ovs_neutron_plugin.ini | grep -v
^# | grep -v ^$
[ovs]
enable_tunneling = True
integration_bridge = br-int
tunnel_bridge = br-tun
local_ip =192.168.5.34
network_vlan_ranges = physnet1,physnet2
bridge_mappings =physnet1:br-ex,physnet2:br-eth2
[agent]
polling_interval = 2
tunnel_types =vxlan
vxlan_udp_port =4789
l2_population = False
arp_responder = False
enable_distributed_routing = False
[securitygroup]
firewall_driver =
neutron.agent.linux.iptables_firewall.OVSHybridIptablesFirewallDriver


[root OScontroller network-scripts]# cat ifcfg-eth0
NAME=eth0
UUID=1b3a9854-df55-43cf-a9b8-21c1a9cc1e5b
DEVICE=eth0
HWADDR=44:1e:a1:75:ea:d6
DEVICETYPE=ovs
OVS_BRIDGE=br-ex
>
TYPE=OVSPort
NM_CONTROLLED=yes


[root OScontroller network-scripts]# cat ifcfg-eth2
NAME=eth2
DEVICE=eth2
HWADDR=44:1e:a1:75:ea:d2
DEVICETYPE=ovs
OVS_BRIDGE=br-eth2
>
TYPE=OVSPort
NM_CONTROLLED=yes


[root OScontroller network-scripts]# cat ifcfg-br-eth2
DEVICE=br-eth2
DEVICETYPE=ovs
TYPE=OVSBridge
BOOTPROTO=static
IPADDR=192.168.31.34
NETMASK=255.255.255.0
>
GATEWAY=192.168.31.1
DNS1=8.8.8.8


[root OScontroller network-scripts]# cat ifcfg-br-ex
DEVICE=br-ex
DEVICETYPE=ovs
TYPE=OVSBridge
BOOTPROTO=static
IPADDR=192.168.5.34
NETMASK=255.255.255.0
>
GATEWAY=192.168.5.1
DNS1=8.8.8.8
DNS2=192.168.5.1


Thank you for your help,
Sara





_______________________________________________
Rdo-list mailing list
Rdo-list redhat com
https://www.redhat.com/mailman/listinfo/rdo-list

To unsubscribe: rdo-list-unsubscribe redhat com


In /etc/neutron/l3_agent.ini, there is a setting which influences the
use of the external bridge:

external_network_bridge =

If this is set to 'br-ex', then the floating IP network is tied to
br-ex, and you can have only one floating IP network.

If on the other hand, this setting is blank, then the floating IP
networks get patched into br-int, and you can have as many as you want.

The bridges that do get used for the floating IP networks should be
included in the bridge mappings.

What I would do in your case is to rename the br-ex bridge to something
else (br-eth1?). Then your bridge mappings becomes:

/etc/neutron/plugins/openvswitch/ovs_neutron_plugin.ini:
bridge_mappings =physnet1:br-eth1,physnet2:br-eth2

You will have to delete br-ex, add br-eth1 (or whatever is
appropriate), and then add the interface to the new bridge. Change the
Neutron plugin files, and restart the Neutron services.


Actually, now that I think about it, you might be able to get away with
leaving the bridge named br-ex, as long as the external_network_bridge
is set to ''. This isn't tested, though, and I wouldn't be terribly
surprised if Neutron didn't like that.

As long as external_network_bridge is set to blank, it doesn't matter how you name your bridges, as long as they are configured correctly in your L2 agent bridge mappings.


We have a combination of command-line parameters and Heat parameters
that will set up the overcloud for multiple floating IP nets at
deployment time. First of all, the external_network_bridge can be set
in the network-environment.yaml file:

parameters:
 Controller-1::NeutronExternalNetworkBridge: "''"

parameter_defaults:
 NeutronExternalNetworkBridge: "''"

Then, in the controller.yaml NIC template, you will want the external
network on a bridge that is named appropriately (not br-ex).

Finally, when you deploy, use the --neutron-bridge-mappings parameter:

openstack overcloud deploy \
--neutron-bridge-mappings physnet1:br-eth1,physnet2:br-eth2 \
-e
/usr/share/openstack-tripleo-heat-templates/environments/network-isolation.yaml
\
-e /home/stack/network-isolation.yaml \
[Other params...]

You will still need to create the floating IP networks after
deployment, we no longer do that at deployment time.

--
Dan Sneddon         |  Principal OpenStack Engineer
dsneddon redhat com |  redhat.com/openstack
650.254.4025        |  dsneddon:irc    dxs:twitter

_______________________________________________
Rdo-list mailing list
Rdo-list redhat com
https://www.redhat.com/mailman/listinfo/rdo-list

To unsubscribe: rdo-list-unsubscribe redhat com

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]