This was the issue that I ran into. I have defined management_network in az. Now it's working fine.

The other issue I am having is I am unable to delete az and its profile while using percona xtradb cluster with pxc_strict_mode enforcing.

openstack loadbalancer availabilityzone delete abc24
(1105, "Percona-XtraDB-Cluster doesn't recommend using SERIALIZABLE isolation with pxc_strict_mode = ENFORCING") (HTTP 500) (Request-ID: req-c40c4902-110d-4b11-8cf8-6542d17768ba)

Ammad

On Mon, Jan 3, 2022 at 12:02 PM Gregory Thiemonge <gthiemonge@redhat.com> wrote:

Hi Ammad,

Do you have a reproducer for this issue (CLI or API calls)?

We recently fixed a similar bug when using Octavia availability zones: https://review.opendev.org/c/openstack/octavia/+/812798
The fix is included in 9.0.1

Greg

On Tue, Dec 28, 2021 at 1:58 PM Ammad Syed <syedammad83@gmail.com> wrote:
Hi,

I am using octavia 9.0. I have created a neutron vlan based management network of octavia. I am creating a loadbalancer with a subnet in the tenant network, its getting created successfully. Then added a listener and created a pool, both created successfully.

Now a weird situation is happening. Now when I add a member in the pool, the management network interface get detached automatically from the AMPHORA instance and amphora keeps in PENDING_UPDATE state.

I have created a dedicated service-project for octavia and management network and subnet resides there.

Any advise how to fix it ?

Ammad