- Users - libvirt List Archives

[libvirt-users] configuring a lxc guest to run 32bits from a 64bits host under libvirt-lxc
by Thierry Parmentelat 14 Jul '14

14 Jul '14

Hi there I have a couple of newbie questions concerning running a 32bits guest in a 64bits host using libvirt - lxc; I’m running libvirt-1.2.5 on fedora18 and fedora20 (a) is this supported at all, and (b) if so how do I configure this for a given guest ? guessing from the mere existence of http://www.redhat.com/archives/libvir-list/2011-February/msg01040.html I would think the answer to (a) is yes; anybody cares to confirm ? As far as (b), I would expect some flag could be set in the XML description of my guest so that libvirt can perform the proper syscall to personality() at the right point in the process hierarchy.. However I have not been able to spot this in the documentation so far, but OTOH the page on the XML configuration here http://libvirt.org/format.html appears to be empty so this might be a temporary hiccup ? in any case, any hint or insight on this matter would be much appreciated — Thierry

2 4

[libvirt-users] libvirt build issue - make check fails , file name long
by Raj Ravi 11 Jul '14

11 Jul '14

Hi, I was trying to build libvirt-0.10.2-29. Now, it builds fine in simple directory setting: like /home/itsme/test/libvirt/ But, if the directory setting is long: like /home/public/test/abc-v1/liibvrt-v1-v2/ = > it fails. Reason for failure is :qemumonitorjsontest I tried running the specific test case like : VIR_TEST_VERBOSE=1 ./qemumonitorjsontest ======================== 1) GetStatus ... libvir: XML-RPC error : Path /home/public/abc_v1_v2/libvirt_0.10.2_29/BUILD/libvirt-0.10.2/tests/qemumonitorjsontest.sock too long for unix socket: File name too long FAILED 2) GetVersion ... libvir: XML-RPC error : Path /home/public/abc_v1_v2/libvirt_0.10.2_29/BUILD/libvirt-0.10.2/tests/qemumonitorjsontest.sock too long for unix socket: File name too long FAILED 3) GetMachines ... libvir: XML-RPC error : Path /home/public/abc_v1_v2/libvirt_0.10.2_29/BUILD/libvirt-0.10.2/tests/qemumonitorjsontest.sock too long for unix socket: File name too long FAILED 4) GetCPUDefinitions ... libvir: XML-RPC error : Path /home/public/abc_v1_v2/libvirt_0.10.2_29/BUILD/libvirt-0.10.2/tests/qemumonitorjsontest.sock too long for unix socket: File name too long FAILED 5) GetCommands ... libvir: XML-RPC error : Path /home/public/abc_v1_v2/libvirt_0.10.2_29/BUILD/libvirt-0.10.2/tests/qemumonitorjsontest.sock too long for unix socket: File name too long FAILED ======================== Is this a known issue ? Is there any specific patch which addresses this issue ? Please let me know. Thanks, Raj

2 1

Re: [libvirt-users] LibVirt + QEMU & no-shutdown
by Michael D 11 Jul '14

11 Jul '14

Scratch that, seems to depend on guest or perhaps even boot device. Finnix live CD halts correctly. Ubuntu 14.04 does not. On 11 July 2014 15:27, Michael D <me(a)md-5.net> wrote: > I am currently running LibVirt with Qemu on Debian Wheezy: > > Versions: > > - libvirtd (libvirt) 1.2.4 > - qemu-x86_64 version 2.0.0 (Debian 2.0.0+dfsg-4~bpo70+1), Copyright > (c) 2003-2008 Fabrice Bellard > > It's all working perfectly, including ACPI: virsh shutdown guest. However > when I issue "halt" from the guest, the guest stops but libvirt thinks it > is still running: (virsh list outputs state: running). virsh dominfo does > the same thing: State: running > > I believe this is due to the -no-shutdown flag passed to qemu, making it > stop the CPUs rather than halt entirely. Is this however a bug that the > status is not updated correctly? I've taken a glance at the source and it > should SIGTERM qemu when it sees the stop, but evidently it isn't. > > > Thanks > > Michael >

1 0

[libvirt-users] LibVirt + QEMU & no-shutdown
by Michael D 11 Jul '14

11 Jul '14

I am currently running LibVirt with Qemu on Debian Wheezy: Versions: - libvirtd (libvirt) 1.2.4 - qemu-x86_64 version 2.0.0 (Debian 2.0.0+dfsg-4~bpo70+1), Copyright (c) 2003-2008 Fabrice Bellard It's all working perfectly, including ACPI: virsh shutdown guest. However when I issue "halt" from the guest, the guest stops but libvirt thinks it is still running: (virsh list outputs state: running). virsh dominfo does the same thing: State: running I believe this is due to the -no-shutdown flag passed to qemu, making it stop the CPUs rather than halt entirely. Is this however a bug that the status is not updated correctly? I've taken a glance at the source and it should SIGTERM qemu when it sees the stop, but evidently it isn't. Thanks Michael

2 2

[libvirt-users] How to config qga to support dompmsuspend
by Hong-Hua.Yin＠freescale.com 10 Jul '14

10 Jul '14

Hi, I tried to run domsuspend command on my PowerPC board but failed. # virsh dompmsuspend sdk --target mem error: Domain sdk could not be suspended error: argument unsupported: QEMU guest agent is not configured It seemed that support suspend-to-mem only from capabilities. # virsh capabilities <capabilities> <host> <uuid>52498f7c-d584-45d8-8a15-345b14fa0d10</uuid> <cpu> <arch>ppc</arch> <topology sockets='1' cores='2' threads='1'/> <pages unit='KiB' size='4'/> <pages unit='KiB' size='1024'/> <pages unit='KiB' size='4096'/> <pages unit='KiB' size='16384'/> <pages unit='KiB' size='65536'/> <pages unit='KiB' size='262144'/> <pages unit='KiB' size='1048576'/> </cpu> <power_management> <suspend_mem/> </power_management> <migration_features> <live/> <uri_transports> <uri_transport>tcp</uri_transport> </uri_transports> </migration_features> <topology> <cells num='1'> <cell id='0'> <memory unit='KiB'>2066304</memory> <cpus num='2'> <cpu id='0' socket_id='0' core_id='0' siblings='0'/> <cpu id='1' socket_id='0' core_id='1' siblings='1'/> </cpus> </cell> </cells> </topology> <secmodel> <model>none</model> <doi>0</doi> </secmodel> <secmodel> <model>dac</model> <doi>0</doi> <baselabel type='kvm'>+0:+0</baselabel> <baselabel type='qemu'>+0:+0</baselabel> </secmodel> </host> <guest> <os_type>hvm</os_type> <arch name='ppc'> <wordsize>32</wordsize> <emulator>/usr/bin/qemu-system-ppc</emulator> <machine maxCpus='1'>g3beige</machine> <machine maxCpus='32'>ppce500</machine> <machine maxCpus='1'>mac99</machine> <machine maxCpus='15'>mpc8544ds</machine> <machine maxCpus='1'>taihu</machine> <machine maxCpus='1'>ref405ep</machine> <machine maxCpus='1'>bamboo</machine> <machine maxCpus='1'>prep</machine> <machine maxCpus='1'>virtex-ml507</machine> <domain type='qemu'> </domain> <domain type='kvm'> <emulator>/usr/bin/qemu-system-ppc</emulator> </domain> </arch> <features> <deviceboot/> <disksnapshot default='on' toggle='no'/> </features> </guest> </capabilities> I read from some document about qemu guest agent but still don't know how to config it as daemon in guest domain. Best Regards, Olivia

3 2

[libvirt-users] snapshot-create error
by Hong-Hua.Yin＠freescale.com 10 Jul '14

10 Jul '14

Hi, What's the difference between QEMU snapshot and libvirt snapshot? # qemu-img create -f qcow2 disk.img 512 Formatting 'disk.img', fmt=qcow2 size=512 encryption=off cluster_size=65536 lazy_refcounts=off # qemu-img info disk.img image: disk.img file format: qcow2 virtual size: 512 (512 bytes) disk size: 136K cluster_size: 65536 Format specific information: compat: 1.1 lazy refcounts: false # /usr/bin/qemu-system-ppc -M ppce500 -enable-kvm -m 512 -nographic -kernel /dev/shm/uImage -initrd /dev/shm/ramdisk -append "root=/dev/ram rw console=ttyS0,115200" -serial tcp::4444,server,telnet -snapshot -hda disk.img QEMU 1.7.0 monitor - type 'help' for more information (qemu) QEMU waiting for connection on: telnet:0.0.0.0:4444,server (qemu) savevm (qemu) info snapshot unknown command: 'info snapshot' (qemu) info snapshots ID TAG VM SIZE DATE VM CLOCK 1 vm-20140709035738 173M 2014-07-09 03:57:38 00:00:07.401 Is there any special qemu command options or kernel configurations to allow libvirt create snapshot? I got the error as below: virsh # snapshot-create snap error: unsupported configuration: internal snapshots and checkpoints require all disks to be selected for snapshot The domain XML file is converted as below steps. # cat >snap.args <<EOF /usr/bin/qemu-system-ppc -name snap -M ppce500 -enable-kvm -m 512 -nographic -kernel /dev/shm/uImage -initrd /dev/shm/ramdisk -append "root=/dev/ram rw console=ttyS0,115200" -serial pty # virsh domxml-from-native qemu-argv snap.args >snap.xml # cat sdk.xml <domain type='kvm'> <name>sdk</name> <memory unit='KiB'>524288</memory> <currentMemory unit='KiB'>524288</currentMemory> <vcpu placement='static'>1</vcpu> <os> <type arch='ppc' machine='ppce500'>hvm</type> <kernel>/dev/shm/uImage</kernel> <initrd>/dev/shm/ramdisk</initrd> <cmdline>root=/dev/ram rw console=ttyS0,115200</cmdline> <boot dev='hd'/> </os> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <emulator>/usr/bin/qemu-system-ppc</emulator> <controller type='usb' index='0'/> <controller type='pci' index='0' model='pci-root'/> <serial type='pty'> <target port='0'/> </serial> <console type='pty'> <target type='serial' port='0'/> </console> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </memballoon> </devices> </domain> Best Regards, Olivia

2 1

[libvirt-users] qemu+tls question
by Fl＠sh 09 Jul '14

09 Jul '14

I`m generate TLS-certificates for my libvirtd and remote, useing http://wiki.libvirt.org/page/TLSSetup . >From console connect to remote host is success, but from virt-manager -- no: Unable to connect to libvirt. Unable to read TLS confirmation: Input/output error Verify that the 'libvirtd' daemon is running on the remote host. Libvirt URI is: qemu+tls://<remote_host>/system Traceback (most recent call last): File "/usr/share/virt-manager/virtManager/connection.py", line 1056, in _open_thread self._backend.open (self._do_creds_password) File "/usr/share/virt-manager/virtinst/connection.py", line 158, in open open_flags) File "/usr/lib64/python2.7/site-packages/libvirt.py", line 102, in openAuth if ret is None:raise libvirtError('virConnectOpenAuth () failed') libvirtError: Unable to read TLS confirmation: Input/output error Any suggestions? -- Fl@sh

1 1

[libvirt-users] Disk and Network Statistics for a Specific Interval
by Sijo Jose 09 Jul '14

09 Jul '14

Hi, virDomainInterfaceStats - This function returns network interface stats for interfaces attached to the domain. virDomainBlockStats- This function returns block device (disk) stats for block devices attached to the domain. Based on which time interval these function returns values. Is it cumulative from the vms uptime..? OR average ...? How can I get the average statistics of Disk and Network for a specific time interval say time x to time y ..? Regards -Sijo

3 2

[libvirt-users] memoryStats question
by Gleb Voronich 09 Jul '14

09 Jul '14

Hello, I'm trying to get memory usage statistic inside a guest machine using virDomainMemoryStats function at libvirt. According to http://libvirt.org/html/libvirt-libvirt.html#virDomainMemoryStatTags the structure could have a lot of useful memory statistic but the amount of parameters depends on the hypervisor and the driver. However using KVM I'm able to get VIR_DOMAIN_MEMORY_STAT_ACTUAL_BALLOON and VIR_DOMAIN_MEMORY_STAT_RSS only. Am I right it is not possible to get additional info such as used/unused memory inside a guest machine in case of KVM?

6 17

[libvirt-users] Libvirt 1.2.6 patch for Hyper-V 2012 R2 ?
by Evgeny Vasilchenko 09 Jul '14

09 Jul '14

Hello, Does anyone know if the below patch for libvirt 1.2.4 has been integrated into the libvirt 1.2.6 source code? http://www.redhat.com/archives/libvir-list/2014-May/msg00509.html https://github.com/vikhyath/libvirt-hyperv-r2-2012 Thank you! Evgeny

2 2