On Tue, Feb 01, 2022 at 08:45:51PM +0100, Martin Perina wrote:

On Tue, Feb 1, 2022 at 8:27 PM Richard W.M. Jones <rjones@redhat.com> wrote:

2022-02-01 19:05:01,952Z ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (EE-ManagedScheduledExecutorService-engineScheduledThreadPool-Thread-19) [330886aa] EVENT_ID: CPU_TYPE_UNSUPPORTED_IN_THIS_CLUSTER_VERSION(156), Host ovirt4410-host moved to Non-Operational state as host CPU type is not supported in this cluster compatibility version or is not supported at all

The host is a nested VM running on old hardware.  I don't care that it's not supported - this is just for testing copying and it'll literally never even need to run a VM.

Is there a way to ignore this and continue?

Hi,

You need to create a custom DC and a cluster with lower compatibility level to be able to use older CPUs.

You can take a look at supported CPUs for each cluster level using ServerCPUList config option:

https://github.com/oVirt/ovirt-engine/blob/master/packaging/dbscripts/upgrad...

OK that operation is underway. I selected CPU type "Nehalem", that being an ancient Intel CPU. The hardware is AMD and there were no AMD options. I'll see if it installs anyway.

What is your hypervisor CPU?

The oVirt host is a VM with an interesting choice of guest CPU now I look at it: processor : 0 vendor_id : AuthenticAMD cpu family : 23 model : 49 model name : AMD EPYC-Rome Processor stepping : 0 microcode : 0x1000065 cpu MHz : 3792.872 cache size : 512 KB physical id : 0 siblings : 1 core id : 0 cpu cores : 1 apicid : 0 initial apicid : 0 fpu : yes fpu_exception : yes cpuid level : 13 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm rep_good nopl cpuid extd_apicid tsc_known_freq pni pclmulqdq ssse3 fma cx16 sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw topoext perfctr_core ssbd ibpb stibp vmmcall fsgsbase tsc_adjust bmi1 avx2 smep bmi2 rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves clzero xsaveerptr wbnoinvd arat umip rdpid arch_capabilities bugs : sysret_ss_attrs spectre_v1 spectre_v2 spec_store_bypass bogomips : 7585.74 TLB size : 1024 4K pages clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: --- and hypervisor CPU: processor : 23 vendor_id : AuthenticAMD cpu family : 23 model : 113 model name : AMD Ryzen 9 3900X 12-Core Processor stepping : 0 microcode : 0x8701013 cpu MHz : 2200.000 cache size : 512 KB physical id : 0 siblings : 24 core id : 14 cpu cores : 12 apicid : 29 initial apicid : 29 fpu : yes fpu_exception : yes cpuid level : 16 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_pstate ssbd mba ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsave_vmload vgif v_spec_ctrl umip rdpid overflow_recov succor smca sme sev sev_es bugs : sysret_ss_attrs spectre_v1 spectre_v2 spec_store_bypass bogomips : 7585.86 TLB size : 3072 4K pages clflush size : 64 cache_alignment : 64 address sizes : 43 bits physical, 48 bits virtual power management: ts ttp tm hwpstate cpb eff_freq_ro [13] [14] Rich.

Regards, Martin

Rich.

-- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/ ~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines.  Supports shell scripting, bindings from many languages.  http://libguestfs.org _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/ community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/ message/NOLWZUDPZMEAJYMNOV75QKQ42BQRQV4B/

-- Martin Perina Manager, Software Engineering Red Hat Czech s.r.o.

-- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-top is 'top' for virtual machines. Tiny program with many powerful monitoring features, net stats, disk stats, logging, etc. http://people.redhat.com/~rjones/virt-top

On Wed, Feb 02, 2022 at 11:07:21AM +0100, Martin Perina wrote:

You cannot mix AMD and Intel processors in a cluster. So if you have an AMD based host, then you need to add it to AMD cluster only

I have two VMs - one for engine and one for node. They are both running on the same physical host (using KVM).

AMD EPYC support is available from 4.3 cluster level, so it should definitely be available in the latest 4.6 CL

I'm more confused here. I'm running what I believe to be the latest oVirt engine (4.4.10.6-1.el8). There are no AMD CPUs offered for cluster CPU type, only Intel CPUs: Intel Nehalem Family Secure Intel Nehalem Family Intel Westmere Family Secure Intel Westmere Family Intel Sandybridge Family (that's the complete list) Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v

Hi, The list should contain more items. Could you please try to create a new cluster using UI, set the compatibility level to 4.6, architecture to x86_64 and check the CPU Type dropdown again? Please try to scroll down the dropdown and if there are no more values than you've already mentioned, please try a different browser to rule out the possibility it is some kind of UI issue. If the UI looks fine, you may try to check the configuration values in the database by running this DB query: select option_value from vdc_options where option_name = 'ServerCPUList' and version = '4.6'; The returned configuration should contain the AMD CPUs. Please, share your findings. Regards, Lucia On Wed, Feb 2, 2022 at 12:30 PM Martin Perina <mperina@redhat.com> wrote:

On Wed, Feb 2, 2022 at 12:11 PM Richard W.M. Jones <rjones@redhat.com> wrote:

...

On Wed, Feb 02, 2022 at 11:07:21AM +0100, Martin Perina wrote:

...

You cannot mix AMD and Intel processors in a cluster. So if you have an AMD based host, then you need to add it to AMD cluster only

I have two VMs - one for engine and one for node. They are both running on the same physical host (using KVM).

...

AMD EPYC support is available from 4.3 cluster level, so it should definitely be available in the latest 4.6 CL

I'm more confused here. I'm running what I believe to be the latest oVirt engine (4.4.10.6-1.el8).

There are no AMD CPUs offered for cluster CPU type, only Intel CPUs:

Intel Nehalem Family Secure Intel Nehalem Family Intel Westmere Family Secure Intel Westmere Family Intel Sandybridge Family

(that's the complete list)

Lucia, any ideas?

...

Rich.

-- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-p2v converts physical machines to virtual machines. Boot with a live CD or over the network (PXE) and turn machines into KVM guests. http://libguestfs.org/virt-v2v

-- Martin Perina Manager, Software Engineering Red Hat Czech s.r.o.

On Wed, Feb 02, 2022 at 01:32:14PM +0100, Lucia Jelinkova wrote:

Hi,

The list should contain more items. Could you please try to create a new cluster using UI, set the compatibility level to 4.6, architecture to x86_64 and check the CPU Type dropdown again?

I created a new DC "Custom", checked compat level is 4.6. Within that, created a new Cluster "Custom", arch is x86_64, and .. you're right! It was a UI issue of some kind because now I see the scrollbar on the right and there are many more machine types. I selected AMD EPYC, so let's see how it goes creating a host. Thanks, Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-df lists disk usage of guests without needing to install any software inside the virtual machine. Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/

Hi! You are running oVirt nested inside KVM? Have you enabled nested virtualization in KVM? Didn't see any mention of that in the thread. That might explain the missing svm flag. https://docs.fedoraproject.org/en-US/quick-docs/using-nested-virtualization-... -Juhani On Wed, Feb 2, 2022 at 3:47 PM Richard W.M. Jones <rjones@redhat.com> wrote:

On Wed, Feb 02, 2022 at 01:06:40PM +0000, Richard W.M. Jones wrote:

...

On Wed, Feb 02, 2022 at 01:32:14PM +0100, Lucia Jelinkova wrote:

...

Hi,

The list should contain more items. Could you please try to create a new cluster using UI, set the compatibility level to 4.6, architecture to x86_64 and check the CPU Type dropdown again?

I created a new DC "Custom", checked compat level is 4.6.

Within that, created a new Cluster "Custom", arch is x86_64, and .. you're right! It was a UI issue of some kind because now I see the scrollbar on the right and there are many more machine types.

I selected AMD EPYC, so let's see how it goes creating a host.

Sadly that didn't fix the problem, I'm still getting:

Host ovirt4410 moved to Non-Operational state as host CPU type is not supported in this cluster compatibility version or is not supported at all

and in engine.log:

2022-02-02 13:35:01,713Z ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (EE-ManagedScheduledExecutorService-engineScheduledThreadPool-Thread-44) [310af602] EVENT_ID: CPU_TYPE_UNSUPPORTED_IN_THIS_CLUSTER_VERSION(156), Host ovirt4410-host moved to Non-Operational state as host CPU type is not supported in this cluster compatibility version or is not supported at all

Mousing over the status, it says it's missing 'model_EPYC' and 'svm' flags. The first one is weird because the emulated model in the guest is:

model name : AMD EPYC-Rome Processor

The second one is correct because I guess KVM is masking this flag from the guest:

flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm rep_good nopl cpuid extd_apicid tsc_known_freq pni pclmulqdq ssse3 fma cx16 sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw topoext perfctr_core ssbd ibpb stibp vmmcall fsgsbase tsc_adjust bmi1 avx2 smep bmi2 rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves clzero xsaveerptr wbnoinvd arat umip rdpid arch_capabilities

Is there not any way to ignore this stuff? I really don't care that this node which is only going to be used for testing won't run VMs optimally. For comparison VMware ESXi 7 also raised minimum CPU types substantially, but offers a simple (unsupported) opt out.

Rich.

-- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/S2Q3M46GUGN5OZ...

Hi! Maybe you have to modify qemu VM CPU settings? I mean that usually if you don't see the svm flag on AMD CPU, virtualization is not available. This page suggests changes to qemu CPU model: https://techviewleo.com/how-to-enable-nested-virtualization-on-kvm-qemu/ I think that the "host-passthrough" setting should let all the CPU flags go through to the VM (https://wiki.openstack.org/wiki/LibvirtXMLCPUModel). If that doesn't do the trick then I'm fresh out of ideas. -Juhani On Wed, Feb 2, 2022 at 4:20 PM Richard W.M. Jones <rjones@redhat.com> wrote:

On Wed, Feb 02, 2022 at 04:09:00PM +0200, Juhani Rautiainen wrote:

...

Hi!

You are running oVirt nested inside KVM? Have you enabled nested virtualization in KVM? Didn't see any mention of that in the thread. That might explain the missing svm flag.

https://docs.fedoraproject.org/en-US/quick-docs/using-nested-virtualization-...

I think so (I thought nested was always enabled on AMD), but let me check ...

This is on the host:

$ cat /sys/module/kvm_amd/parameters/nested 1

I believe that means yes.

Inside the guest (the oVirt node):

$ cat /sys/module/kvm_amd/parameters/nested cat: /sys/module/kvm_amd/parameters/nested: No such file or directory

but I believe that's correct since we're not trying to do double nesting with KVM. TBH I don't care at all if the VMs run slowly (with TCG) or even not at all. This is just for testing. I just want a way to ignore the error and bring up the host.

Rich.

...

-Juhani

On Wed, Feb 2, 2022 at 3:47 PM Richard W.M. Jones <rjones@redhat.com> wrote:

...

On Wed, Feb 02, 2022 at 01:06:40PM +0000, Richard W.M. Jones wrote:

...

On Wed, Feb 02, 2022 at 01:32:14PM +0100, Lucia Jelinkova wrote:

...

Hi,

The list should contain more items. Could you please try to create a new cluster using UI, set the compatibility level to 4.6, architecture to x86_64 and check the CPU Type dropdown again?

I created a new DC "Custom", checked compat level is 4.6.

Within that, created a new Cluster "Custom", arch is x86_64, and .. you're right! It was a UI issue of some kind because now I see the scrollbar on the right and there are many more machine types.

I selected AMD EPYC, so let's see how it goes creating a host.

Sadly that didn't fix the problem, I'm still getting:

Host ovirt4410 moved to Non-Operational state as host CPU type is not supported in this cluster compatibility version or is not supported at all

and in engine.log:

2022-02-02 13:35:01,713Z ERROR [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector] (EE-ManagedScheduledExecutorService-engineScheduledThreadPool-Thread-44) [310af602] EVENT_ID: CPU_TYPE_UNSUPPORTED_IN_THIS_CLUSTER_VERSION(156), Host ovirt4410-host moved to Non-Operational state as host CPU type is not supported in this cluster compatibility version or is not supported at all

Mousing over the status, it says it's missing 'model_EPYC' and 'svm' flags. The first one is weird because the emulated model in the guest is:

model name : AMD EPYC-Rome Processor

The second one is correct because I guess KVM is masking this flag from the guest:

flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm rep_good nopl cpuid extd_apicid tsc_known_freq pni pclmulqdq ssse3 fma cx16 sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw topoext perfctr_core ssbd ibpb stibp vmmcall fsgsbase tsc_adjust bmi1 avx2 smep bmi2 rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves clzero xsaveerptr wbnoinvd arat umip rdpid arch_capabilities

Is there not any way to ignore this stuff? I really don't care that this node which is only going to be used for testing won't run VMs optimally. For comparison VMware ESXi 7 also raised minimum CPU types substantially, but offers a simple (unsupported) opt out.

Rich.

-- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/S2Q3M46GUGN5OZ...

-- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-builder quickly builds VMs from scratch http://libguestfs.org/virt-builder.1.html

Host-passthrough & nested virtualization are mandatory. Best Regards,Strahil Nikolov On Wed, Feb 2, 2022 at 18:54, Richard W.M. Jones<rjones@redhat.com> wrote: On Wed, Feb 02, 2022 at 05:42:33PM +0200, Juhani Rautiainen wrote:

Hi!

Maybe you have to modify qemu VM CPU settings? I mean that usually if you don't see the svm flag on AMD CPU, virtualization is not available. This page suggests changes to qemu CPU model: https://techviewleo.com/how-to-enable-nested-virtualization-on-kvm-qemu/

I think that the "host-passthrough" setting should let all the CPU flags go through to the VM (https://wiki.openstack.org/wiki/LibvirtXMLCPUModel).

Yes, thank you, you are absolutely right. I had the CPU model set to "host-model" (where libvirt tries to create a model approximating the host).  Changing it to "host-passthrough" (ie. same as qemu -cpu host) fixed it and allowed me to install oVirt node in the guest and activate it. Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-df lists disk usage of guests without needing to install any software inside the virtual machine.  Supports Linux and Windows. http://people.redhat.com/~rjones/virt-df/ _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/W4434OIVOGS2KC...

https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/8/htm... Best Regards,Strahil Nikolov On Thu, Feb 3, 2022 at 10:53, Richard W.M. Jones<rjones@redhat.com> wrote: On Thu, Feb 03, 2022 at 06:08:55AM +0000, Strahil Nikolov wrote:

Host-passthrough & nested virtualization are mandatory.

Yes indeed.  I don't know if I missed this in the docs, but it wasn't obvious.  However using host-passthrough fixed it. Rich.

Best Regards, Strahil Nikolov

    On Wed, Feb 2, 2022 at 18:54, Richard W.M. Jones     <rjones@redhat.com> wrote:     On Wed, Feb 02, 2022 at 05:42:33PM +0200, Juhani Rautiainen wrote:     > Hi!     >     > Maybe you have to modify qemu VM CPU settings? I mean that usually if     > you don't see the svm flag on AMD CPU, virtualization is not     > available. This page suggests changes to qemu CPU model:     > https://techviewleo.com/how-to-enable-nested-virtualization-on-kvm-qemu/     >     > I think that the "host-passthrough" setting should let all the CPU     > flags go through to the VM     > (https://wiki.openstack.org/wiki/LibvirtXMLCPUModel).

    Yes, thank you, you are absolutely right.

    I had the CPU model set to "host-model" (where libvirt tries to create     a model approximating the host).  Changing it to "host-passthrough"     (ie. same as qemu -cpu host) fixed it and allowed me to install oVirt     node in the guest and activate it.

    Rich.

    --     Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/     ~rjones     Read my programming and virtualization blog: http://rwmj.wordpress.com     virt-df lists disk usage of guests without needing to install any     software inside the virtual machine.  Supports Linux and Windows.     http://people.redhat.com/~rjones/virt-df/         _______________________________________________     Users mailing list -- users@ovirt.org     To unsubscribe send an email to users-leave@ovirt.org     Privacy Statement: https://www.ovirt.org/privacy-policy.html     oVirt Code of Conduct: https://www.ovirt.org/community/about/     community-guidelines/     List Archives:     https://lists.ovirt.org/archives/list/users@ovirt.org/message/     W4434OIVOGS2KCGVISROD6Y7JPLP6MNS/    

-- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com virt-builder quickly builds VMs from scratch http://libguestfs.org/virt-builder.1.html _______________________________________________ Users mailing list -- users@ovirt.org To unsubscribe send an email to users-leave@ovirt.org Privacy Statement: https://www.ovirt.org/privacy-policy.html oVirt Code of Conduct: https://www.ovirt.org/community/about/community-guidelines/ List Archives: https://lists.ovirt.org/archives/list/users@ovirt.org/message/FIR2NPU2FVCNMX...