- Users - libvirt List Archives

[libvirt-users] libvirt does not show same CPU Model as /proc/cpuinfo for CPU Model info.
by akhilesh rawat 31 Jan '17

31 Jan '17

Hi , Created new thread . Environment: Bare Metal server + CentOs with qemu/KVM +libvirt for virtualization Guest Instantiated with virt-install with forced CPU model like below virt-install --virt-type kvm --name compute-0 --cpu Haswell,+fma,+movbe,+fsgsbase,+bmi1,+hle,+avx2,+smep,+bmi2,+erms,+invpcid,+rtm --ram=61440 --vcpus=20 --os-type=linux --os-variant=generic After guest installation /proc/cpuinfo show model name as Haswell . However Libvirt virsh capabilities show CPU configuration as "SandyBridge . " 1. Could I get explanation of the why inconsistency . 2. Is this expected behaviour 3. Why libvirt does not show same CPU Model as /proc/cpuinfo for CPU Model info. My Aim : Nested Virtualization and Nested Guest needs CPU model as Haswell strictly . 4. IS it possible alternative to succeed in my aim ? Thanks a lot for support. ! Guest CPU config: processor : 19 vendor_id : GenuineIntel cpu family : 6 model : 60 model name : Intel Core Processor (Haswell) stepping : 1 microcode : 0x1 cpu MHz : 1995.144 cache size : 4096 KB fpu : yes fpu_exception : yes cpuid level : 13 wp : yes flags : fpu de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx rdtscp lm constant_tsc rep_good nopl eagerfpu pni pclmulqdq ssse3 cx16 pcid sse4_1 sse4_2 x2apic popcnt tsc_deadline_timer aes xsave avx hypervisor lahf_lm xsaveopt bogomips : 4105.33 clflush size : 64 cache_alignment : 64 address sizes : 46 bits physical, 48 bits virtual power management: <capabilities> <host> <uuid>95c4e625-3383-4ec1-9ee4-c47a87d93b97</uuid> <cpu> <arch>x86_64</arch> <model>SandyBridge</model> <vendor>Intel</vendor> <topology sockets='20' cores='1' threads='1'/> <feature name='hypervisor'/> <feature name='osxsave'/> <feature name='pcid'/> <pages unit='KiB' size='4'/> <pages unit='KiB' size='2048'/> Host CPU config: processor : 23 vendor_id : GenuineIntel cpu family : 6 model : 45 model name : Intel(R) Xeon(R) CPU E5-2620 0 @ 2.00GHz stepping : 7 microcode : 0x710 cpu MHz : 1200.000 cache size : 15360 KB physical id : 1 siblings : 12 core id : 5 cpu cores : 6 apicid : 43 initial apicid : 43 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 dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm pcid dca sse4_1 sse4_2 x2apic popcnt tsc_deadline_timer aes xsave avx lahf_lm ida arat epb pln pts dtherm tpr_shadow vnmi flexpriority ept vpid xsaveopt bogomips : 4004.01 clflush size : 64 cache_alignment : 64 address sizes : 46 bits physical, 48 bits virtual power management: Br Aki

2 2

[libvirt-users] LibVirt query CPU Model support and restore operation
by akhilesh rawat 27 Jan '17

27 Jan '17

Hello , It was working good working with kvm management tools using libvirt . virsh/virt-manager . But then i got annoyed when management tool did not allow me to change the CPU model while creating new virtual machine . error root@kvm-server qemu]# virt-install --virt-type kvm --name compute-2 --cpu Haswell-noTSX --ram=61440 --vcpus=20 --os-type=linux --os-variant=generic --disk compute2-disk0.qcow2,device=disk,bus=ide,size=300,sparse=true,format=qcow2 --pxe --network bridge=virbr0,model=e1000 --network bridge=virbr0,model=e1000 --network bridge=virbr0,model=virtio --network bridge=virbr0,model=virtio --graphics vnc,port=5906 --noautoconsole Starting install... ERROR unsupported configuration: guest and host CPU are not compatible: Host CPU does not provide required features: invpcid, erms, bmi2, smep, avx2, bmi1, fsgsbase, movbe, fma Domain installation does not appear to have been successful. I could not resolve this error . Finally i came across the post which say that libvirt cross verifies flags under CPU model present in /usr/share/libvirt/cpu_map.xml and cat /proc/cpuinfo . An yes the flag complained were not present in host proc/cpuinfo. Question 1: Is it really needed for Libvirt to do this check . AS KVM seems to allow support Haswell with native commands . Why libvirt is doing this check then ? As i was not sure where the problem lies and i was using libvirt 2.0.0 version i upgraded to 2.5.0 by compiling form source . But then i faces issue using virsh/virt-manager as seem it was not compatilble with upgarded version of libvirt. Now my system is little messed around libvirt . I did reinstall/reboot Quite many things not working . Question 2 : How can i restore original Libvirt 2.0.0 . br aki

3 8

[libvirt-users] Unresticted guest trunk network interface
by Dennis Jacobfeuerborn 26 Jan '17

26 Jan '17

Hi, is there way to create a network interface for a guest that just forwards packets tagged with a vlan id? Currrently I have a guest that has configured 10+ interfaces and for each one I add I have to create a new vlan interface and a bridge on the host which makes things rather cumbersome. What I'd like to have is just one interface eth0 in the guest and then be able to create eth0.2, eth0.3 in the guest and have the host just pass them through without interference. I found some guides on the internet but these always come with some limitation such as that if you have a vlan interface on the host like e.g. eth0.8 then the guest no longer receives packets in vlan 8 as there apparently get consumed by the host. Is there any solution for this? Regards, Dennis

1 0

[libvirt-users] KVM live migration issues for Windows guests
by Fabrizio Soppelsa 23 Jan '17

23 Jan '17

Hi, We're repeatedly facing a live migration issue with Windows guests, would be great if someone could send their thoughts/suggestions/experience on how to further troubleshoot this. When we live migrate a Windows instance, it gets migrated (guest is running on the destination host), but it eventually internally hangs. From the Windows logs, the operating system has tried to shut down during the live migration. There is no info at all, even at debug level, in libvirt logs. I'm thinking to go with perf kvm or maybe kernel upgrade, but it might be I'm missing something obvious in Windows for this scenario? QEMU 2.0.0 Libvirtd 1.2.2 Host kernel version : 3.13.0-40-generic Windows version: Windows 2012 R2 Has RBD volumes attached driver name='qemu' type='raw' cache='writeback'/, but reproduced also without. Thanks! ![](https://link.nylas.com/open/bi1c6xnw46zc0gy4ozrqerfkk/local-e0bf2fff- 21e6?r=bGlidmlydC11c2Vyc0ByZWRoYXQuY29t)

1 0

[libvirt-users] gluster store and autostart - but fails
by lejeczek 23 Jan '17

23 Jan '17

hi all I've a few guest which work/run perfectly fine, I believe, except for autostart. Configuration of system, gluster and libvirt is pretty regular and not complex. Errors I see: ... failed to initialize gluster connection (src=0x7f9424266350 priv=0x7f94242922b0): Transport endpoint is internal error: Failed to autostart VM 'rhel-work2': failed to initialize gluster connection (src=0x7f9 failed to initialize gluster connection (src=0x7f942423fef0 priv=0x7f9424256320): Transport endpoint is internal error: Failed to autostart VM 'rhel-work3': failed to initialize gluster connection (src=0x7f9 failed to initialize gluster connection (src=0x7f9424261b20 priv=0x7f94242a18b0): Transport endpoint is internal error: Failed to autostart VM 'rhel-work1': failed to initialize gluster connection (src=0x7f9 ... I tried to make systemd libvirtd to wait for gluster: After=glusterd.service but if that's all required then, well, still fails. Would you have any suggestions? Many thanks, L.

2 1

[libvirt-users] Controlling the name of the 'tap0' device, in a bridged networking setup
by Govert 21 Jan '17

21 Jan '17

Hi, I'm trying to control the name of the 'tap0' device that gets created as I start a domain that uses bridged networking. The XML specification of the domain contains the following configuration <interface type='bridge'> <source bridge='br0'/> </interface> The libvirt documentation ( http://libvirt.org/formatdomain.html#elementsNICSBridge) and other discussions online tell me that I just need to include the <target dev='desired_dev_name'/> tag in the XML specification of the domain under the <interface> tag. Unfortunately doing so appears to have no effect; the tun device created and 'enslaved' in the bridge is still called 'tap0'. Interestingly, I never get a tun device with a name prefixed by 'vnet' or 'vif' which, according to the documentation, is the default behaviour (?). The host is running CentOS 7, and virsh is used to start the domain. Best, Govert

2 4

[libvirt-users] io=native & io=threads
by W Kern 17 Jan '17

17 Jan '17

Googling provides lots of interesting info on the use of these in various situations such as SSD, number of VMs in the pool etc. What is the default in Libvirt (or is the default 'neither') Sincerely W Kern

2 2

[libvirt-users] Regarding Migration Statistics
by Anubhav Guleria 14 Jan '17

14 Jan '17

Greetings, I am writing a code using libvirt API to migrate VM between two physical hosts *(QEMU/KVM) , *say some *n *number of times. *1)* I am using right now* virDomainPtr virDomainMigrate (.......) *and to calculate the total migration time I am using something like this: *clock_gettime(CLOCK_MONOTONIC_RAW,&begin); * *migrate*(domainToMigrate,nodeToMigrate); *clock_gettime(CLOCK_MONOTONIC_RAW,&end);* *Total Migration Time = end.tv_sec-begin.tv_sec* Is this correct way to calculate total migration time. And is there some way to calculate the downtime (not how to set it)? *2) *I am interested in identifying in particular other statistics of migration like : *Number of iterations in Pre Copy*, *Memory transferred in each iteration* etc. I was going through the API and found* virDomainJobInfo <http://libvirt.org/html/libvirt-libvirt-domain.html#virDomainJobInfo> and virDomainGetJobStats <http://libvirt.org/html/libvirt-libvirt-domain.html#virDomainGetJobStats> *functions.But how to use them is not very clear. Can anyone point me to right place to achieve this objective? Thanks in advance. And sorry if that was too silly to ask. Anubhav

3 3

[libvirt-users] changing to cache='none' on the fly?
by W Kern 10 Jan '17

10 Jan '17

As is already well documented, we find that we need cache='none' to support migration, otherwise there is the chance of a hang and/or failure to pivot. However we prefer the default of cache=writethrough when operating in production. Our practice is to 'shutdown' the VM completely make the change with virsh edit, then restart. Then we have to repeat the process to revert back once we migrate. Is it possible to change that function on the fly and avoid the shutdown/start process? Note: we have had inconsistent results with virsh edit first and then a reboot. A complete shutdown seems to be necessary, but I am hoping there is some other procedure available. Our images are all qcow2 (1.1) Sincerely, W Kern PixelGate Networks

2 1

[libvirt-users] 回复： [libvirt] xml config nested
by 放牛班的春天 10 Jan '17

10 Jan '17

thanks your mail，now below： If the KVM supports nested VMX and QEMU starts with the arguments -enable-kvm and -cpu ..., + vmx, then the LOCK bit of the guest MSR_IA32_FEATURE_CONTROL and Enable VMX out of SMX operation bit will be set. Well, I should add -enable-kvm parameters of the document which xml position, Exactly speaking, how to amend xml configuration file? ------------------ 原始邮件 ------------------ 发件人: "Daniel P. Berrange";<berrange(a)redhat.com>; 发送时间: 2017年1月5日(星期四) 晚上8:53 收件人: "放牛班的春天"<446844717(a)qq.com>; 抄送: "libvirt-users"<libvirt-users(a)redhat.com>; 主题: Re: [libvirt] xml config nested NB, in future please don't CC all possible mailing lists at once. Just pick the most appropriate mailing list for your question. I've dropped libvirt-list & libvirt-announce from the CC, since this is a question most suited for libvirt-users. On Thu, Jan 05, 2017 at 11:44:29AM +0800, 放牛班的春天 wrote: > How to configure libvirt, so qemu-kvm support nested virtualization, virtual machine installed operating system is windows_7_ultimate_sp1_x64_dvd_618537.iso, configure libvirt xml file is as follows: > > <cpu mode='custom' match='exact'> > <model fallback='allow'>core2duo</model> > <feature policy='require' name='vmx'/> > </cpu> > > > or > > <cpu mode='host-model'> > <model fallback='allow'/> > </cpu> > > or > <cpu mode='host-passthrough'> > <topology sockets='2' cores='2' threads='2'/> > </cpu> Yes, that's the key guest configuration step - exposing the 'vmx' feature to the guest. In addition to that though, you need to make sure the kvm-intel kernel module in your host has the "nested=1" parameter set. eg in /etc/modprobe.d/kvm.conf you want options kvm_intel nested=1 If loaded correctly you should see # cat /sys/module/kvm_intel/parameters/nested Y if it says "N", then nested VMX will be disabled. Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://entangle-photo.org -o- http://search.cpan.org/~danberr/ :|

1 0