Hello,

We are deploying RHOSO 18 for a customer and we have an issue while provisionning edpm nodes.

After creating the BMH and OpenStackDataPlaneNodeSet (osdpns), the nodes get provisioned and marked so on BMH level. the status of osdpns is also OK (waiting for for deployment)
but the nodes are actually stuck not getting booted with an error of `/dec/disk/by-label/img-rootfs does not exist` before entering emergency shell.

the nodes after applying osdpns.
 
 

oc get bmh 
NAME                 STATE         CONSUMER                     ONLINE   ERROR   AGE 
vel1-nfv12-kvm0106   provisioned   openstack-data-plane-vradi   true             54m vel1-nfv12-kvm0107   provisioned   openstack-data-plane-vradi   true             54m

 
 
the osdpns status
 
 

oc get osdpns
NAME                         STATUS   MESSAGE 
openstack-data-plane-vradi   False    NodeSet setup ready, waiting for Deployment...

after a launch of deployment.
the ansiblee job stuck waiting for connection which did not get enabled as the node is on dracut emergency mode..

oc logs -l app=openstackansibleee -f --max-log-requests 10 
Skipping callback 'minimal', as we already have a stdout callback. Skipping callback 'oneline', as we already have a stdout callback.   PLAYBOOK: bootstrap.yml ******************************************************** 
2 plays in /usr/share/ansible/collections/ansible_collections/osp/edpm/playbooks/bootstrap.yml   PLAY [Check remote connections] ************************************************   TASK [Wait for connection] ***************************************************** 
task path: /usr/share/ansible/collections/ansible_collections/osp/edpm/playbooks/bootstrap.yml:7

#To Reproduce:

Not sure if the issue can be reproduced.

#Expected behavior

The expected behavior is to have the node provisioned and started then receive the configurations and openstack pods after a OpenStackDataPlaneDeployment is launched.

#Screenshots

Screenshot for the device path error on the node console.

#Device Info:

Hardware Specs: Cisco UCS B200 M6 2 Socket Blade Server
                Disks are exposed to the nodes over SAN and multipath is used.
OS Version: The image set on BMH level is rhel9.4 exposed on a local webserver and the information regarding this image are specified on the BMH level.

Bug impact

We have a RHOSO deployment which is blocked for the customer.
 

Additional context
the image : edpm-hardened-uefi.qcow2 is used even if we have specified other image information on the bmh file.

• We have both tested specifying a rootDeviceHints to /dev/dm-0 and also with id wwn.

  rootDeviceHints:
       wwn: "0x600a098038314648475d5231626f5067"
   #   deviceName: "/dev/dm-0"

• regarding the image

  we have tested without specifying image param on the bmh file and also with speciying ``url: 'http://<local repos ip>/rhel-9.4-x86_64-kvm.qcow2'`` : as our environment is a disconnected env the edpm-hardened-uefi.qcow2 is used by default.
 
  we tried configuring the osdpns osImage but the image is not downloaded (no url attribute to configure local repos so by giving only the image name it is not found while provisionnig )
 
 

  baremetalSetTemplate:
     deploymentSSHSecret: dataplane-ansible-ssh-private-key-secret
     bmhNamespace: openstack
     cloudUserName: ansible
   # osImage: rhel-9.4-x86_64-kvm.qcow2
     bmhLabelSelector:
       app: openstack
       workload: compute
       vnf: vradi
     ctlplaneInterface: eno10.1023
     dnsSearchDomains:
       - <domain>

• we tried to pass the kernel param on the osdpns file ``edpm_kernel_args: "rd.multipath=default"`` as suggested on other discussion but didnot resolve the issue.