Remove NVME-related UUID from images `/usr/etc/nvme/hostid` `/usr/etc/nvme/hostnqn`

[travier]

See: #1445 (comment)

There are apparently hardcoded NMVE ids in the images right now and we should remove those to make sure those are unique and generated per-node.

[travier]

See https://issues.redhat.com/browse/OCPBUGS-34629

[openshift-bot]

Issues go stale after 90d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.
Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle stale

[openshift-bot]

Stale issues rot after 30d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.
Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle rotten
/remove-lifecycle stale

[jlebon]

/remove-lifecycle rotten

[dustymabe]

In Fedora and EL10 the %post that generates this was removed:

• https://src.fedoraproject.org/rpms/nvme-cli/c/629f632c97ab888f2ae3a8bc3eb2bacf0959dd2d
• https://gitlab.com/redhat/centos-stream/rpms/nvme-cli/-/commit/762906d9286ed3d83af70ff6d620ba32b5addd3e
• https://issues.redhat.com/browse/RHEL-68495

I'm not sure if there are supporting changes made in nvme-cli and other packages that made this OK in EL10+.

I do know if I implement this change to rm -fv /etc/nvme/hostid /etc/nvme/hostnqn then the isci tests we have still work.

[dustymabe]

I guess either way we don't want the same ID baked in so what we have right now is clearly wrong.

We can start with deleting them and then if we need to generate them on first boot for EL9 we can create a service to do that.