I don't use Kolla, but the management network config in Trove is:

[DEFAULT]

management_networks = 

---

Lingxian Kong

Senior Cloud Engineer (Catalyst Cloud)

Trove Project Lead (OpenStack)

OpenStack Cloud Provider Project Lead (Kubernetes)

On Wed, Oct 20, 2021 at 9:23 PM wodel youchi <wodel.youchi@gmail.com> wrote:

Hi Lingxian, and thanks.

Could you be more specific? What is this management network? in globals.yml this what I have in terme of networking :

kolla_internal_vip_address: "10.10.3.1"
kolla_internal_fqdn: "dashinternal.domain.tld"
kolla_external_vip_address: "x.x.x.x"
kolla_external_fqdn: "dash.domain.tld"
network_interface: "bond0"
kolla_external_vip_interface: "bond1"
api_interface: "bond1.30"
storage_interface: "bond1.10"
tunnel_interface: "bond1.40"
octavia_network_interface: "{{ api_interface }}"
neutron_external_interface: "bond2"
neutron_plugin_agent: "openvswitch"

What is this management network? Do I have to create it? If yes how?

Regards.

Le mer. 20 oct. 2021 à 00:48, Lingxian Kong <anlin.kong@gmail.com> a écrit :

Hi Wodel,

There is a management network (Neutron network) configured for communication between controller services and guest agent, you need to config in the infra router (using the network vlan) layer to make sure they can talk.

---

Lingxian Kong

Senior Cloud Engineer (Catalyst Cloud)

Trove Project Lead (OpenStack)

OpenStack Cloud Provider Project Lead (Kubernetes)

On Wed, Oct 20, 2021 at 10:30 AM wodel youchi <wodel.youchi@gmail.com> wrote:

Hi,

I am trying to deploy Trove. I am using the Kolla-ansible and Openstack wallaby version.

From the documentation, the Trove guest agent, which runs inside the Tove instance communicates with the trove-taskmanager via rabbitmq, how is this done?

The rabbitmq is running in the api network, the instance is running in the tunnel (tenant) network, and in my case, those networks are in different vlans, how should I configure this?

Regards.