We are trying to deploy OCP on Baremetal servers (Lenovo servers SR630 and SR65) using Assited installer Saas mode. dhcp hosts (Master and Worker nodes) connect fine during discovery but don't get networking after rebooting and fail the pull ignition from the bootstrap. the interfaces came up fine during discovery, but not the second boot. this happen only with Lenovo servers.

NAME="Red Hat Enterprise Linux CoreOS"
ID="rhcos"
ID_LIKE="rhel fedora"
VERSION="412.86.202301061548-0"
VERSION_ID="4.12"
PLATFORM_ID="platform:el8"

Configure dnsmasq to provide dns and dhcp service in lab for servers and also NFS server to boot servers on iso discovery image. Create OCP cluster using Assisted installer on Red Hat Hybrid Cloud console. Create iso image (full media version) and boot servers using this iso image through NFS server installed in Lab.

1. Configure dnsmasq and NFS servers 
2. Create OCP cluster on Red Hat Hybrid Cloud console and create discovery iso image
3. Boot servers using NFS server 

servers connect fine during discovery but don't get networking after rebooting and fail the pull ignition from the bootstrap. the interfaces came up fine during discovery, but not the second boot.

servers connect fine during discovery, get networking after rebooting and the pull ignition from the bootstrap to continue OCP installation.

I was able to manually configure networking in an emergency shell and get the install to continue. The Workaround consist to enter in emergency shell during second boot and manually configure server network interface.

To enter in emergency shell, press e during boot and add the parameter below on kernel command

rd.break=initqueue


Network config with ip tool as nmcli not installed yet at this boot stage (initqueue)

ip link set ens4f0 up
ip addr add 192.168.27.20/25 dev ens4f0