- Users - libvirt List Archives

[libvirt-users] Strange IPv6 Issues
by Tim Schumacher 23 Jun '14

23 Jun '14

Hi folks, I'm having the problem that my guests are most of the time not reachable via IPv6 and cant connect to IPv6 hosts, but if I ping the public IPv6-Adress of my bridge, the host can use IPv6 again. This is for incomming and outgoing traffic. I've attached my network xml file. virsh version reports this: Compiled against library: libvirt 1.2.4 Using library: libvirt 1.2.4 Using API: QEMU 1.2.4 Running hypervisor: QEMU 2.0.0 Routing on the guest looks like this: $ ip -6 route 2a01:4f8:200:2265:5054:ff:febe:e8e6 dev eth0 proto kernel metric 256 fe80::/64 dev eth0 proto kernel metric 256 default via fe80::5054:ff:fe32:e58 dev eth0 metric 1024 Routing on the host looks like this: $ ip -6 route 2a01:4f8:200:2265::/64 dev eth0 proto kernel metric 256 2a01:4f8:200:2265::/64 dev virbr0 proto kernel metric 256 fe80::/64 dev eth0 proto kernel metric 256 fe80::/64 dev vnet0 proto kernel metric 256 fe80::/64 dev vnet1 proto kernel metric 256 fe80::/64 dev virbr0 proto kernel metric 256 fe80::/64 dev vnet2 proto kernel metric 256 fe80::/64 dev vnet3 proto kernel metric 256 fe80::/64 dev vnet4 proto kernel metric 256 fe80::/64 dev vnet5 proto kernel metric 256 default via fe80::1 dev eth0 metric 1024 I'm running Debian Jessy on the host and on the guest. That means that they all are running Kernel 3.14. If you have any idea what could be causing this or if you need any information please post. Thanks for your time and help Tim

1 0

[libvirt-users] ERROR: Domain not found: no domain with matching name 'ubuntu'
by Sanket Panhale 22 Jun '14

22 Jun '14

Hi guys, I am new to QEMU-KVM, libvmi and libvirt stuff. Libvmi uses libvirt. I am trying to to run process-list example of libvmi and getting error as below. It seems that this error may be due to libvirt as it is not able to find domain. I seek your kind help on below error: spanhal1@seclab2:~/KVMModule/libvmi-0.10.1$ sudo ./examples/process-list ubuntu libvir: QEMU error : Domain not found: no domain with matching name 'ubuntu' VMI_ERROR: Failed to get domain id from name. Failed to init LibVMI library.1 Here are few details of step I performed: 1. I have installed KVM module. (version 3.5) 2. Installed QEMU-KVM. (version 1.2.0) 3. Installed libvirt(version 0.1.2.5) 4. Installed libvirt-bin(version 0.9.8-2ubuntu17.19) 5. Installed libvmi. (version 0.10.1) 6. Then I am trying to run process-list example and I am getting error as above. sudo ./examples/process-list ubuntu To deal with this, I tried creating VM with domain name ubuntu using steps given at below link. https://help.ubuntu.com/community/KVM/CreateGuests http://virt-tools.org/learning/install-with-command-line/ Create VM: $sudo apt-get install virtinst $sudo fallocate -l 8192M /var/lib/libvirt/images/guest.img $sudo qemu-img create -f qcow2 /var/lib/libvirt/images/guest.qcow2 8192 spanhal1@seclab2:~/KVMModule$ sudo virt-install --connect qemu:///system -n ubuntu -r 512 -f guest.qcow2 -s 12 -c /home/spanhal1/KVMModule/ubuntu-12.04.4-desktop-amd64.iso --vnc --noautoconsole --os-type linux --accelerate --network=network:default ERROR Host does not support any virtualization options Here are few workaround and information for this error: 1. spanhal1@seclab2:~/KVMModule$ modprobe kvm spanhal1@seclab2:~/KVMModule$ lsmod | grep kvm kvm_intel 137888 0 kvm 422160 1 kvm_intel 2. spanhal1@seclab2:~/KVMModule$ kvm-ok INFO: /dev/kvm exists KVM acceleration can be used Check if virutalization is enabled by BIOS: spanhal1@seclab2:/var/log/libvirt$ sudo modprobe msr spanhal1@seclab2:/var/log/libvirt$ sudo rdmsr 0x3a 5 (Values 3 and 5 mean it's activated). Libvirt log: 2014-06-12 20:06:24.033+0000: 1305: error : virCommandWait:2188 : internal error Child process (LC_ALL=C PATH=/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/sbin:/sbin:/bin /usr/local/bin/qemu-system-x86_64 -help) status unexpected: exit status 1 2014-06-12 20:06:24.034+0000: 1305: error : virCommandWait:2188 : internal error Child process (LC_ALL=C PATH=/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/sbin:/sbin:/bin /usr/local/bin/qemu-system-x86_64 -help) status unexpected: exit status 1 Machine Details: Intel(R) Core(TM) i3-3240 CPU @ 3.40GHz 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 rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm pcid sse4_1 sse4_2 popcnt tsc_deadline_timer xsave avx f16c lahf_lm arat epb xsaveopt pln pts dtherm tpr_shadow vnmi flexpriority ept vpid fsgsbase smep erms OS: Ubuntu 12.04.1 LTS Config file for Libvmi: ubuntu { ostype = "Linux"; sysmap = "/boot/System.map-3.5.0-22-generic"; linux_name = 0x460; linux_tasks = 0x240; linux_mm = 0x278; linux_pid = 0x2b4; linux_pgd = 0x48; } Thanks, Sanket.

3 5

[libvirt-users] /etc/machine-id
by Gene Czarcinski 21 Jun '14

21 Jun '14

I was under the impression that the value of /etc/machine-id was unqiue for every OS installation on a system. This seems to be the case on real hardware. That is, if I install Fedora 19, Fedora 20, and Fedora-rawhide on a system, they will each have a different value for /etc/machine-id. This does not seem to be the case for qemy-kvm-libvirt virtual systems. SOmetimes they are different but I just noticed that a virtual system with both Fedora 20 and Fedora rawhide installed (btrfs partitioning), the two installations have the same value in /etc/machine-id. The two installs have different ext4 partitions for /boot, different subvols for the respective rootfs and the same/shared subvol for /home. Comment? Bug? Gene

2 1

[libvirt-users] Managed USB passthrough
by Michael Han 21 Jun '14

21 Jun '14

Does the "managed" attribute of the hostdev element for a USB device make a difference? The Domain XML Format docs has the following sentence for the hostdev element: If managed is omitted or "no", and for USB devices, the user is responsible to call virNodeDeviceDettach (or virsh nodedev-dettach) before starting the guest or hot-plugging the device, and virNodeDeviceReAttach (or virsh nodedev-reattach) after hot-unplug or stopping the guest. Is that still accurate? The device seems to get passed through fine to my LXC container without any dettach/reattach calls. But I'm seeing some other strange USB behaviour in the guest and just want to see if it's at all related to the managed attribute. Cheers.

1 0

[libvirt-users] Access permissions of passed through character devices
by Pickartz, Simon 20 Jun '14

20 Jun '14

Dear all, I am using libvirt in conjunction with LXC. In accordance with http://libvirt.org/formatdomain.html#elementsHostDevCaps I pass an InfiniBand adapter through to the guest container. However, within the container the access permissions are not the same. Within the host they are like this: crw-rw-rw- 1 root root 10, 56 20. Jun 09:40 rdma_cm crw-rw-rw- 1 root root 231, 224 20. Jun 09:41 ucm0 crw-rw-rw- 1 root root 231, 0 20. Jun 09:41 umad0 crw-rw-rw- 1 root root 231, 1 20. Jun 09:41 umad1 crw-rw-rw- 1 root root 231, 192 20. Jun 09:41 uverbs0 Within the guest, only root may read or write from/to the device: crw------- 1 root root 10, 56 20. Jun 09:40 rdma_cm crw------- 1 root root 231, 224 20. Jun 09:41 ucm0 crw------- 1 root root 231, 0 20. Jun 09:41 umad0 crw------- 1 root root 231, 1 20. Jun 09:41 umad1 crw------- 1 root root 231, 192 20. Jun 09:41 uverbs0 Do you have any ideas how I can tell lxc via libvirt to change the permissions in such a way that they correspond to those on the host system? Best regards, Simon

1 0

[libvirt-users] converted VMDK disk iamge and Virtio driver
by Paul Raines 19 Jun '14

19 Jun '14

I have a CentOS 5 VM in vmware I wanted to move to work on KVM/libvirt. I used qemu-img to convert it to a qcow2 image, loaded virt-manager and created a new VM pointed at that qcow2 image. The boot paniced when it tried to mount the root filesystem also complaining about not being able to remount the swap partition. I played around and was able to get it to boot by changing the Virtual disk bus to IDE from Virtio. Searched about Virtio and I got to http://wiki.libvirt.org/page/Virtio and ran the mkinitrd mentioned to force loading of the virtio_pci and virtio_blk drivers. But upon shutting down, configuring the bus back to Virtio and booting it panics in the exact same way. I tried again after adding virtio_blk into /etc/modprobe.conf as scsi_hostadapter and rerunning mkinitrd but it still fails. Any clues what I am missing? --------------------------------------------------------------- Paul Raines http://help.nmr.mgh.harvard.edu MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging 149 (2301) 13th Street Charlestown, MA 02129 USA The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.

2 3

[libvirt-users] bridge could not be initialized
by abhishek jain 19 Jun '14

19 Jun '14

Hi I'm trying to run VM using libvirt.xml file and getting following error... virsh start instance-00000003 error: Failed to start domain instance-00000003 error: internal error Process exited while reading console log output: failed to launch bridge helper kvm: -netdev bridge,br=qbr1f2191ce-38,id=hostnet0: Device 'bridge' could not be initialized Below is the libvirt.xml file.. domain type="kvm"> <uuid>d9f007b9-584f-41c8-a706-e096f354575b</uuid> <name>instance-00000003</name> <memory>524288</memory> <vcpu>1</vcpu> <os> <type>hvm</type> <kernel>/opt/stack/data/nova/instances/d9f007b9-584f-41c8-a706-e096f354575b/kernel</kernel> <initrd>/opt/stack/data/nova/instances/d9f007b9-584f-41c8-a706-e096f354575b/ramdisk</initrd> <cmdline>root=/dev/vda console=tty0 console=ttyS0</cmdline> </os> <features> <acpi/> <apic/> </features> <clock offset="utc"> <timer name="pit" tickpolicy="delay"/> <timer name="rtc" tickpolicy="catchup"/> </clock> <devices> <disk type="file" device="disk"> <driver name="qemu" type="qcow2" cache="none"/> <source file="/opt/stack/data/nova/instances/d9f007b9-584f-41c8-a706-e096f354575b/disk"/> <target bus="virtio" dev="vda"/> </disk> <interface type="bridge"> <mac address="fa:16:3e:26:55:ff"/> <model type="virtio"/> <source bridge="qbr1f2191ce-38"/> <target dev="tap1f2191ce-38"/> <filterref filter="nova-instance-instance-00000003-fa163e2655ff"/> </interface> <serial type="file"> <source path="/opt/stack/data/nova/instances/d9f007b9-584f-41c8-a706-e096f354575b/console.log"/> </serial> <serial type="pty"/> </devices> </domain> Please help regarding this. Thanks Abhishek Jain

3 2

[libvirt-users] dumpxml removes <source> portion from network descritiption
by Thomas Stein 19 Jun '14

19 Jun '14

Hello. I have a strange error with one of my VMs. If i use dumpxml to save a VM description the source part of the network part is missing. This prevents redifining the machine from the dumped xml file: This is how it looks before dumpxml: <interface type='bridge'> <mac address='52:54:00:34:a1:56'/> <source bridge='br406'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> And afterwards: <interface type='bridge'> <mac address='52:54:00:34:a1:56'/> <target dev='vnet6'/> <model type='virtio'/> <alias name='net0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> Someone has an idea whats causing this? Is there a special option maybe? thanks and best regards t.

2 2

[libvirt-users] Use keepalive to monitor health of qemu-kvm process
by Abhijeet Rastogi 19 Jun '14

19 Jun '14

Hi everyone, I've had instances where qemu-kvm process dies and become a zombie. At that state, libvirtd reports the instance to be running and I can't restart the VM because libvirtd fails to kill the qemu-kvm process. Finding the root cause of that is a topic for another day. I also see this line in logs: Jun 15 13:36:57 d12566 libvirtd: 10424: warning : virKeepAliveTimerInternal:156 : No response from client 0x1600420 after 5 keepalive messages in 31 seconds Is it possible to monitor the keepAlive thing via command-line ? (as a nagios check). I know that I can poll for that line in the logs but is there a better way? I couldn't find any option in virsh to do a KeepAlive check. Thanks -- Cheers, Abhijeet Rastogi (shadyabhi)

1 0

[libvirt-users] Live migrate failure
by Tuân Tạ Bá 18 Jun '14

18 Jun '14

Hi everyone, I use KVM on Openstack Cloud. And KVM migrated successfull, but not ping. *Source Host SIDE: *"/var/log/libvirt/libvirtd.log" 2014-06-16 05:15:50.599+0000: 2096: warning : qemuMigrationCancelDriveMirror:1383 : Unable to stop block job on drive-virtio-disk0 2014-06-16 05:15:50.601+0000: 2096: *warning : qemuMigrationCancelDriveMirror:1383 : Unable to stop block job on drive-virtio-disk1* *2014-06-16 05:15:50.651+0000: 2096: error : virCommandWait:2348 : internal error: Child process (ovs-vsctl --timeout=5 get Interface tap6977e4a1-cd external_ids:PortData) unexpected exit status 1: ovs-vsctl: no key "PortData" in Interface record "tap6977e4a1-cd" column external_ids* 2014-06-16 05:15:50.651+0000: 2096: error : virNetDevOpenvswitchGetMigrateData:219 : Unable to run command to get OVS port data for interface tap6977e4a1-cd: Operation not permitted 2014-06-16 05:15:50.651+0000: 2096: error : qemuMigrationCookieNetworkAlloc:367 : Unable to run command to get OVS port data for interface tap6977e4a1-cd: Operation not permitted 2014-06-16 05:15:50.651+0000: 2096: warning : qemuMigrationRun:3262 : Unable to encode migration cookie 2014-06-16 05:15:53.321+0000: 2096: warning : virAuditSend:135 : Failed to send audit message virt=kvm op=stop reason=migrated vm="instance-0001b37c" uuid=631ac4c7-8e84-4cf6-ad98-93877cdbe1bd vm-pid=-1: Operation not permitted *Destination Host SIDE:* "/var/log/libvirt/libvirtd.log" ...... 2014-06-16 05:15:41.419+0000: 2229: warning : virAuditSend:135 : Failed to send audit message virt=kvm resrc=cgroup reason=allow vm="instance-000 1b37c" uuid=631ac4c7-8e84-4cf6-ad98-93877cdbe1bd cgroup="/sys/fs/cgroup/devices/machine/instance-0001b37c.libvirt-qemu/" class=major category=pty maj=88 acl=rw: Operation not permitted 2014-06-16 05:15:41.620+0000: 2229: warning : qemuDomainObjEnterMonitorInternal:1175 : This thread seems to be the async job owner; entering moni tor without asking for a nested job is dangerous 2014-06-16 05:15:43.552+0000: 2229: warning : virAuditSend:135 : Failed to send audit message virt=kvm resrc=vcpu reason=start vm="instance-0001b 37c" uuid=631ac4c7-8e84-4cf6-ad98-93877cdbe1bd old-vcpu=0 new-vcpu=2: Operation not permitted 2014-06-16 05:15:43.552+0000: 2229: warning : virAuditSend:135 :* Failed to send audit message *virt=kvm op=start reason=migrated vm="instance-0001 b37c" uuid=631ac4c7-8e84-4cf6-ad98-93877cdbe1bd vm-pid=21063:* Operation not permitted* 1) I'm using libvirt version 1.1.1, libvirt-bin 1.1.1-0ubuntu8.5~cloud0 libvirt0 1.1.1-0ubuntu8.5~cloud0 python-libvirt 1.1.1-0ubuntu8.5~cloud0 2) #cat /etc/libvirt/qemu.conf ... user = "root" group = "root" I searched GOOGLE, René speaks "downgrading to libvirt 1.1.0 and reinstalling everything. They will ok". But I don't think that. How to resolve this error?. Thanks all! -- TuanTaBa

2 2