In Octavia, pool members are not necessarily a part of the OpenStack environment, so may not have compute IDs at all. When creating pool members, you will note, we do not ask for a compute ID. Pool members are only IP addresses and ports. If a member is a nova instance, searching the server list by IP address is your best way of finding the compute ID. I have not seen a script that automates this. Michael On Mon, Dec 8, 2025 at 8:47 AM Eugen Block <eblock@nde.ag> wrote:

It should probably be:

openstack loadbalancer amphora list --long

Zitat von Eugen Block <eblock@nde.ag>:

...

Hi,

I currently don’t have access to environments using Octavia, but I believe ‚openstack amphora list --long‘ (or similar, typing on my phone) contains the compute_id. Does that command help?

Regards, Eugen

Zitat von Max Harmathy <maxh@uhurutec.com>:

...

Hello OpenStack community,

I regularly face the following problem: From time to time I get notified about issues with some load-balancers where often the issue lies in connectivity between the health monitor and some of the pool members.

Now for investigating the problem I often need to know the ID of the server (compute id) for a certain pool member. I haven't found a straightforward way, since the information the OpenStack client provides for a given pool member (openstack loadbalancer member show ...) doesn't include the compute id.

I usually try to find out the IP address of the pool member and then match that against the list of servers. But that process is quite tedious.

Is there a better way to do this natively or has someone made a tool for that job?

Regards, Max

-- Max Harmathy UhuruTec AG

Tel +49 151 16 341 719 max.harmathy@uhurutec.com https://www.uhurutec.com

Industriestraße 4 D-70565 Stuttgart Vorsitzender des Aufsichtsrats: Maximilian Buchberger Vorstand: Matthias Haag, Jan Klippel Registergericht Stuttgart, HRB 785603 USt-IdNr: DE354021144