Created attachment 1263190 [details] win10-cpu-rhv4.1 Description of problem: The processor status of win10 guest is abnormal in rhv4.1 Version-Release number of selected component (if applicable): rhv:4.1.1.2-0.1.el7 How reproducible: 100% Steps to Reproduce: 1.Using virt-v2v convert win10 guest from kvm to rhv 2.Import the guest from export domain to data domain after finishing v2v conversion 3.Power on the guest and check cpu status in device manager, the cpu status shows as below, pls refer to screenshot "win10-cpu-rhv4.1" "Currently this hardware device is not connected to the computer (Code 45). To fix this problem, reconnect the hardware device to the computer." Actual results: As above description Expected results: The processor status of win10 guest is normal in rhv4.1 Additional info: 1.Also could reproduce the problem when import win10 guest from vmware at rhv4.1 2.Can't reproduce the problem on win7,win2008r2 and win8 guests
What's the qemu command line? Does it happen when you run win10 guest regularly from RHV, without the conversion? (I assume it's either a conversion - v2v issue - or a qemu-kvm issue - need to understand where the bug is).
(In reply to Yaniv Kaul from comment #1) > What's the qemu command line? > Does it happen when you run win10 guest regularly from RHV, without the > conversion? > > (I assume it's either a conversion - v2v issue - or a qemu-kvm issue - need > to understand where the bug is). If import the win10 guest from vmware directly without v2v conversion on rhv4.1, the cpu status is abnormal If import the win8 guest from vmware directly without v2v conversion on rhv4.1, the cpu status is normal, pls refer to screenshot"CPU-win10-win8" So the bug is related to rhv itself,not v2v, thanks
Created attachment 1263280 [details] CPU-win10-win8
By the way, this bug was confirmed by Shahar at https://bugzilla.redhat.com/show_bug.cgi?id=1372668#c14
I have a feeling this is not a bug. https://bugzilla.redhat.com/show_bug.cgi?id=1372668#c13 describes what is actually happening here. Unless oVirt has a way to express that we want a Sandybridge CPU ...
(In reply to Richard W.M. Jones from comment #5) > I have a feeling this is not a bug. > > https://bugzilla.redhat.com/show_bug.cgi?id=1372668#c13 > > describes what is actually happening here. Unless oVirt has a > way to express that we want a Sandybridge CPU ... Tomas?
A follow up to my comment 5 is that in very recent virt-v2v (not RHEL 7.4, but maybe 7.5) it is now possible to pass through CPU vendor, model and topology from source to destination: https://github.com/libguestfs/libguestfs/commit/7f940c3e3a8de21c27e517b4ccde02fa7e7b287f Only the topology is passed through to oVirt right now, because I could not work out how to pass the CPU vendor and/or model. I checked the ovirt-engine code and I'm fairly sure it's not possible, but maybe I have missed something there. BTW if you are going to implement this in oVirt, please don't invent your own lists of CPU vendors / models. Libvirt already has a taxonomy for this, see the output of `virsh capabilities' and the file `/usr/share/libvirt/cpu_map.xml'. Another aspect of this is getting the CPU vendor and model from VMware. AFAICT VMware doesn't expose it, but maybe we could look harder for it there.
It is not possible so far. We can add the support once the support on v2v will arrive. Setting this BZ to future so far and in meanwhile opened a new doc issue to have it documented as a limitation: https://bugzilla.redhat.com/show_bug.cgi?id=1437064
The processor status of win2016 guest also is abnormal in rhv4.1, pls refer to screenshot"win2016-cpu"
Created attachment 1274811 [details] win2016-cpu
The processor status of win10 guest also is abnormal after converting from vmware to openstack by virt-v2v, pls refer to screenshot"win10-openstack" The openstack host's cpu is Intel(R) Xeon(R) CPU E5-2407 0 @ 2.20GHz Packages:virt-v2v-1.36.3-5.el7.x86_64/libguestfs-1.36.3-5.el7.x86_64
Created attachment 1290470 [details] win10-openstack
Michal any chance we can implement the changes done in BZ 1437064 into RHV (mainly exposing and being able to set the CPU model and vendor from VMWare.
The CPU is abnormal in win2016 guest on rhv4.2.7.5-0.1.el7ev, pls refer to screenshot "win2016-rhv4.2", but CPU is normal in win10 and win2016 guests on rhv4.2
Created attachment 1512962 [details] win2016-rhv4.2
This bug didn't get any attention for a while, we didn't have the capacity to make any progress. If you deeply care about it or want to work on it please assign/target accordingly
ok, closing. Please reopen if still relevant/you want to work on it.