Josh, can you describe the motivation in more detail? You can encounter two MAC addresses in one deployment, I don't see a reason why to prevent from that. If the host will be provisioned in a different subnet, than this constraint could be problem.

Not sure about serial number, not all servers provide those.

This bug will be part of Satellite NEXT (stable is 6.2 at the moment).

Moving this bug to POST for triage into Satellite 6 since the upstream issue http://projects.theforeman.org/issues/16890 has been resolved.

1439850

Verified on Satellite 6.3 Snap 20. 

Using the foreman-discovery-image-3.1.1-22, doing PXEless discovery:

1. Have some previously registered (content) host
2. Boot a virtual system using the foreman-discovery image, go through the discovery process, and check the Hosts->Discovered hosts table. The system that was just discovered should be present.
3. Just to be sure, modify the virtual system HW (add some memory or virtual HDD), and try to re-run the discovery procedure. The discovered system should be updated, not duplicated.
4. Remove the discovered host.
5. Edit previously registered host, and change its MAC address to MAC address of the virtual system. 
6. Try to re-run the discovery process on the virtual system. Instead of the success message, the process will loop back to the start of the IP address configuration wizard. There is the following message in /var/log/foreman/production.log:

Host discovery failed, facts: {"lib"=>"/usr/share/fdi/facts:/opt/extension/facts", "virtual"=>"kvm", "is_virtual"=>true, "architecture"=>"x86_64", "hardwaremodel"=>"x86_64", "os"=>{"name"=>"RedHat"
, "family"=>"RedHat", "release"=>{"major"=>"7", "minor"=>"3", "full"=>"7.3"}}, "network_ens9"=>"192.168.122.0", "network_lo"=>"127.0.0.0", "hardwareisa"=>"x86_64", "physicalprocessorcount"=>1, "netmask"=>"255.255.255.0", "memorysize"=>"
992.57 MB", "memorysize_mb"=>"992.57", "ps"=>"ps -ef", "fqdn"=>"fdi", "uniqueid"=>"a8c07b7a", "dhcp_servers"=>{"system"=>"192.168.122.1", "ens9"=>"192.168.122.1"}, "hostname"=>"fdi", "macaddress"=>"52:54:00:b4:c4:03", "bios_vendor"=>"Se
aBIOS", "bios_version"=>"1.10.2-1.fc26", "bios_release_date"=>"04/01/2014", "manufacturer"=>"QEMU", "productname"=>"Standard PC (i440FX + PIIX, 1996)", "serialnumber"=>"Not Specified", "uuid"=>"9C322EB1-C718-4816-98C2-11C81247857A", "ty
pe"=>"Other", "selinux"=>false, "interfaces"=>"ens9,lo", "ipaddress_ens9"=>"192.168.122.123", "macaddress_ens9"=>"52:54:00:b4:c4:03", "netmask_ens9"=>"255.255.255.0", "mtu_ens9"=>1500, "ipaddress_lo"=>"127.0.0.1", "netmask_lo"=>"255.0.0
.0", "mtu_lo"=>65536, "gid"=>"root", "id"=>"root", "ipaddress"=>"192.168.122.123", "processors"=>{"models"=>["Intel Core Processor (Haswell, no TSX)"], "count"=>1, "physicalcount"=>1}, "processor0"=>"Intel Core Processor (Haswell, no TS
X)", "processorcount"=>1, "facterversion"=>"2.4.6", "blockdevice_sr0_size"=>218103808, "blockdevice_sr0_vendor"=>"QEMU", "blockdevice_sr0_model"=>"QEMU DVD-ROM", "blockdevice_vda_size"=>3221225472, "blockdevice_vda_vendor"=>"0x1af4", "b
lockdevices"=>"sr0,vda", "discovery_version"=>"3.1.1", "discovery_release"=>"20170323.1", "discovery_bootif"=>"52:54:00:b4:c4:03", "discovery_bootip"=>"192.168.122.123", "speed_ens9"=>"100", "duplex_ens9"=>"full", "port_ens9"=>"MII", "a
uto_negotiation_ens9"=>"true", "wol_ens9"=>false, "link_ens9"=>"true", "link_lo"=>"true", "discovery_kexec"=>"kexec-tools 2.0.7 released 05 September 2016\n", "discovery_ip_cidr"=>"192.168.122.123/24", "discovery_ip"=>"192.168.122.123",
 "discovery_netmask"=>"255.255.255.0", "discovery_gateway"=>"192.168.122.1", "discovery_dns"=>"192.168.122.1", ""=>""}

7. Change the MAC address of the previously registered host to the original value.
8. Re-run the discovery process on the virtual system. It will be successful this time.

VERIFIED

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2018:0336