Bug 1949413

Description of problem:
The boot is always done by adding "2" to the boot device. For example it creates the string "sda2". See the code here: https://github.com/openshift/assisted-installer/blob/749f7737317f407c97d6ff347c569aac71d224db/src/ops/ops.go#L225-L229

However, when using by-path devices, this logic is incorrect. We see errors in the log:

level=info msg="Done writing image to disk"
level=info msg="Setting efibootmgr to boot from disk"
level=info msg="mount.nfs: Failed to resolve server /dev/disk/by-path/pci-0000: Name or service not known\n"
level=info msg="failed executing nsenter [-t 1 -m -i -- mount /dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:0:02 /mnt], env vars [PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin TERM=xterm USER_UID=0 HTTPS_PROXY= http_proxy= HTTP_PROXY= https_proxy= container=podman no_proxy= NO_PROXY= PULL_SECRET_TOKEN=b3BlbnNoaWZ0LXJlbGVhc2UtZGV2K29jbV9hY2Nlc3NfM2ZlN2U2NTRiN2Q3NDk2ZThmMjA3MTNlMjk2ZTlmZDc6VlA0SzZaNkZWS0lWRVBYQ1Y4VFk3NjFTWjFSSk01Mkk5OEU2UTRYQ1c3WlRaMzZVQ0RSWTlVTkVMMDQ3QUxJVA== HOSTNAME=master-0-1 HOME=/root], error exit status 32, waitStatus 32, Output \"mount.nfs: Failed to resolve server /dev/disk/by-path/pci-0000: Name or service not known\""
level=error msg="Failed to mount device /dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:0:02, err: failed executing nsenter [-t 1 -m -i -- mount /dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:0:02 /mnt], Error exit status 32, LastOutput \"mount.nfs: Failed to resolve server /dev/disk/by-path/pci-0000: Name or service not known\""
level=warning msg="Failed to set boot order" error="failed executing nsenter [-t 1 -m -i -- mount /dev/disk/by-path/pci-0000:03:00.0-scsi-0:0:0:02 /mnt], Error exit status 32, LastOutput \"mount.nfs: Failed to resolve server /dev/disk/by-path/pci-0000: Name or service not known\""


How reproducible:
On some multi-disk hosts


Steps to Reproduce:
1. Create a libvirt environment of 3 masters and 3 workers
2. The workers in my case had 2 disks, sda and sdb, of identical sizes
3. Deploy a cluster


Actual results:
Automatic boot order is not set correctly, and the user has to fix it manually. In my case, I didn't get a chance to fix the boot order manually and the installation failed - we are still investigating if it's due to other errors.


Expected results:
We shouldn't see these errors in the log. The user should not have to fix the boot order manually.