Open Stack

Tested on 3 nodes and 1 compute node, it is working fine by adding the
eth14 to the br-lbaas bridge.
Satish your solution will save 1 physical interface, will try your solution
using VLAN later.

Thanks




On Thu, Oct 1, 2020 at 10:27 PM pwm <pwm2012 at gmail.com> wrote:

> ok, will try to test it on 3 nodes this weekend. I see you are using the
> Linux bridge, thanks for sharing your solution.
>
>
> On Tue, Sep 29, 2020 at 10:01 PM Satish Patel <satish.txt at gmail.com>
> wrote:
>
>> Perfect,
>>
>> Now try same way on 3 nodes and tell me how it goes?
>>
>> Because I was having issue in my environment to make it work so I used
>> different method which is here on my blog
>> https://satishdotpatel.github.io//openstack-ansible-octavia/
>>
>> Sent from my iPhone
>>
>> On Sep 29, 2020, at 8:27 AM, pwm <pwm2012 at gmail.com> wrote:
>>
>> 
>> Hi,
>>  I'm testing on AIO before moving to a 3 nodes controller. Haven't tested
>> on 3 nodes controller yet but I do think it will get the same issue.
>>
>> On Tue, Sep 29, 2020 at 3:43 AM Satish Patel <satish.txt at gmail.com>
>> wrote:
>>
>>> Hi,
>>>
>>> I was dealing with the same issue a few weeks back so curious are you
>>> having this problem on AIO or 3 node controllers?
>>>
>>> ~S
>>>
>>> On Sun, Sep 27, 2020 at 10:27 AM pwm <pwm2012 at gmail.com> wrote:
>>> >
>>> > Hi,
>>> > I using the following setup for Octavia load balancer on OVS
>>> > Ansible openstack_user_config.yml
>>> >     - network:
>>> >         container_bridge: "br-lbaas"
>>> >         container_type: "veth"
>>> >         container_interface: "eth14"
>>> >         host_bind_override: "eth14"
>>> >         ip_from_q: "octavia"
>>> >         type: "flat"
>>> >         net_name: "octavia"
>>> >         group_binds:
>>> >           - neutron_openvswitch_agent
>>> >           - octavia-worker
>>> >           - octavia-housekeeping
>>> >           - octavia-health-manager
>>> >
>>> > user_variables.yml
>>> > octavia_provider_network_name: "octavia"
>>> > octavia_provider_network_type: flat
>>> > octavia_neutron_management_network_name: lbaas-mgmt
>>> >
>>> > /etc/netplan/50-cloud-init.yaml
>>> >  br-lbaas:
>>> >             dhcp4: no
>>> >             interfaces: [ bond10 ]
>>> >             addresses: []
>>> >             parameters:
>>> >                 stp: false
>>> >                 forward-delay: 0
>>> >  bond10:
>>> >             dhcp4: no
>>> >             addresses: []
>>> >             interfaces: [ens16]
>>> >             parameters:
>>> >                 mode: balance-tlb
>>> >
>>> > brctl show
>>> > bridge name     bridge id                         STP enabled
>>>  interfaces
>>> > br-lbaas            8000.d60e4e80f672       no
>>> 2ea34552_eth14
>>> >
>>>                    bond10
>>> >
>>> > However, I am getting the following error when creating the load
>>> balance
>>> > octavia.amphorae.drivers.haproxy.rest_api_driver [-] Could not connect
>>> to instance. Retrying.: requests.exceptions.ConnectionError:
>>> HTTPSConnectionPool(host='172.29.233.47', port=9443
>>> >
>>> > The Octavia api container unable to connect to the amphora instance.
>>> > Any missing configuration, cause I need to manually add in the eth14
>>> interface to the br-lbaas bridge in order to fix the connection issue
>>> > brctl addif br-lbaas eth14
>>> >
>>> > Thanks
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-discuss/attachments/20201002/b24ce43f/attachment.html>