Hi all, I am currently experiencing issues while setting up a basic ironic deployment. My environment consists of a controller running Ubuntu 24.04 and a baremetal host with IPMI support but no operating system installed. I am using kolla-ansible for the Flamingo version OpenStack deployment, and all configuration is being managed through the globals.yml file. One of the controller’s network interfaces is configured to provide DHCP and is connected to a physical switch. I have attached an image (Figure 1) illustrating the physical layout of my setup. My controller had the following configuration: { Two interface (eth0 and eth1) eth0: 10.42.0.1 with netmask 24 eth1: No IP assignment } My compute had the following configuration: { 5 interfaces (eth0, eth1, eth2, eth3, IPMI) eth1: No IP assignment and is connected to switch (MAC: 6F:F1:34:E9:01:29) IPMI: 10.42.0.100 connected to switch } In my globals.yml file, I have the following configuration: { openstack_release: "2025.2" # Flamingo kolla_internal_vip_address: "10.42.0.1" docker_configure_for_zun: true containerd_configure_for_zun: true containerd_grpc_gid: 42463 enable_openstack_core: true enable_haproxy: false enable_cinder: false enable_etcd: true enable_fluentd: true enable_ironic: true enable_kuryr: true enable_magnum: true enable_zun: true ironic_dnsmasq_dhcp_ranges: - range: "10.42.0.5,10.42.0.200" ironic_dnsmasq_boot_file: "pxelinux.0" ironic_cleaning_network: "public1" ironic_dnsmasq_default_gateway: "10.42.0.1" ironic_inspector_kernel_cmdline_extras: ['ipa-lldp-timeout=90.0', 'ipa-collect-lldp=1'] } After completing the deployment, I verified that all the required services are running as docker containers. Immediately afterward, I executed the necessary command to generate the admin credentials, which allows me to use the OpenStack client. 1. $pip install python-openstackclient -c https://releases.openstack.org/constraints/upper/master 2. $kolla-ansible post-deploy 3. $/path/to/venv/share/kolla-ansible/init-runonce After running the command, the following baseline was created. 1. Two network (public1 and demo-net), where public is a flat network 2. cirros image file 3. flavors After that, I uploaded a working kernel, initramfs, and ISO image to OpenStack. I verified that these images were functioning correctly by successfully launching an instance. Next, during the node enrollment process, I ran the following command to enroll my baremetal node. 1. $openstack flavor create --ram 32768 --vcpu 16 --disk 120 baremetal 2. $openstack flavor set --property resources:CUSTOM_IRONIC=1 baremetal 3. $openstack flavor set --property resources:VCPU=0 baremetal 4. $openstack flavor set --property resources:MEMORY_MB=0 baremetal 5. $openstack flavor set --property resources:DISK_GB=0 baremetal 6. $openstack baremetal node create --driver ipmi --deploy-interface direct --driver-info ipmi_username=<IPMI_usernmae> --driver-info ipmi_password=<IPMI_password> --driver-info ipmi_address=10.42.0.100 --driver-info deploy_kernel=<kernel uploaded> --driver-info deploy_ramdisk=<initramfs uploaded> --property local_gb=20 --resource-class baremetal --name My_node After enrolling the node as described above, and since I do not have an ironic Inspector setup, I manually created a port for the enrolled node using the MAC address 6F:F1:34:E9:01:29, which corresponds to the LAN interface. At this point, I am unsure how to proceed with booting the desired image onto the bare-metal host. I would greatly appreciate any guidance on the next steps required to deploy the image to the machine. Best, Vincent
