- Users - libvirt List Archives

[libvirt-users] Problems running virtual machines after an upgrade
by Bhasker C V 29 Feb '12

29 Feb '12

Hi all, I upgraded qemu-kvm and libvirt to the new versions qemu-kvm at : 1.0 libvirt at : git version 0.9.10 I lost the original machines but the xml files were are all present (it is not showing up in virsh list --all) But when I import an xml I get an error: Attached is some information which might be helpful Can someone help please ? If this is a compatibility issue please can someone tell me which is the latest version of qemu-kvm and libvirt combination which will work ? I cannot use qemu plain since it does not yet support boot from scsi thanks ----------------------------------------------- $ sudo virsh list --all Id Name State ---------------------------------------------------- $ lsmod | grep kvm kvm_intel 112920 0 kvm 280656 1 kvm_intel $ /usr/local/sbin/libvirtd --version /usr/local/sbin/libvirtd (libvirt) 0.9.10 $ ./qemu-system-x86_64 --version QEMU emulator version 1.0-bcv (qemu-kvm-1.0), Copyright (c) 2003-2008 Fabrice Bellard $ sudo virsh define ./debian.xml error: Failed to define domain from ./debian.xml error: unknown OS type hvm $ cat debian.xml <domain type='kvm'> <name>debian</name> <uuid>0aab8869-ff87-14ce-a00b-944cdca7acf7</uuid> <memory>524288</memory> <currentMemory>524288</currentMemory> <vcpu cpuset='6-7'>2</vcpu> <cputune> <vcpupin vcpu='0' cpuset='6'/> <vcpupin vcpu='1' cpuset='7'/> </cputune> <os> <type arch='x86_64' machine='pc-0.14'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> <pae/> </features> <cpu match='exact'> <model>core2duo</model> <vendor>Intel</vendor> <topology sockets='1' cores='2' threads='1'/> <feature policy='disable' name='lahf_lm'/> </cpu> <clock offset='localtime'> <timer name='hpet' tickpolicy='catchup'/> </clock> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/local/bin/kvm</emulator> <disk type='file' device='disk'> <driver name='qemu' type='qcow2'/> <source file='/VIRT/disks/DEBIAN'/> <target dev='sda' bus='scsi'/> <address type='drive' controller='0' bus='0' unit='0'/> </disk> <disk type='block' device='cdrom'> <driver name='qemu' type='raw'/> <target dev='hdc' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' unit='0'/> </disk> <controller type='scsi' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </controller> <controller type='ide' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <interface type='bridge'> <mac address='52:54:00:71:62:f3'/> <source bridge='virt'/> <model type='e1000'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <serial type='pty'> <target port='0'/> </serial> <serial type='pty'> <target port='0'/> </serial> <console type='pty'> <target type='serial' port='0'/> </console> <input type='mouse' bus='ps2'/> <graphics type='vnc' port='-1' autoport='yes' keymap='en-gb'/> <sound model='ich6'> <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x0'/> </sound> <video> <model type='cirrus' vram='9216' heads='1'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </memballoon> </devices> </domain> -- Bhasker C V

3 3

[libvirt-users] Virtualizing a an old SuSE system
by Dale Amon 29 Feb '12

29 Feb '12

I'm in the midst of an attempt to convert an old and rather large SuSE server (5 disks) into a virsh loadable VM. Has anyone else dealt with the issues of systems of this sort? I'm at the moment trying to hand construct a machine xml file for it. I managed to create one which would load but not start. I'd appreciate some pointers and can supply whatever info you'd like about what I am trying to do and what the dd'd image of the system disk and other disks are like. Not that they do have LVM's, although the boot should be off a base partition that is not in the LVM.

3 6

[libvirt-users] Fwd: Unable to connect qemu:///system on Fedora16 X86_64
by suyi wang 29 Feb '12

29 Feb '12

---------- 已转发邮件 ---------- 发件人： suyi wang <wangsuyi640(a)gmail.com> 日期： 2012年2月29日上午10:57 主题： Re: [libvirt-users] Unable to connect qemu:///system on Fedora16 X86_64 收件人： Eric Blake <eblake(a)redhat.com> So, thanks a lot . Eric. I attached the figure，which showed when I ran the command "virt-manager". I also tested the commad as follows, sa you can see my output. [root@localhost ~]# virt-manager --debug 2012-02-29 10:56:13,555 (cli:71): virt-manager startup 2012-02-29 10:56:13,555 (virt-manager:292): Launched as: /usr/share/virt-manager/virt-manager.py --debug 2012-02-29 10:56:13,555 (virt-manager:293): GTK version: (2, 24, 7) 2012-02-29 10:56:13,555 (virt-manager:294): virtManager import: <module 'virtManager' from '/usr/share/virt-manager/virtManager/__init__.pyc'> 2012-02-29 10:56:13,671 (engine:555): No inspection thread because libguestfs is too old, not available, or libvirt is not thread safe. 2012-02-29 10:56:13,674 (engine:346): About to connect to uris ['qemu:///system'] 2012-02-29 10:56:13,720 (engine:471): window counter incremented to 1 2012-02-29 10:56:13,726 (connection:954): Scheduling background open thread for qemu:///system 2012-02-29 10:56:13,727 (connection:1140): Background 'open connection' thread is running Thanks! Suyi Wang 2012/2/29 Eric Blake <eblake(a)redhat.com> > On 02/28/2012 10:23 AM, suyi wang wrote: > > Dear all: > > I ust installed Fedora16Fedora16 X86_64 on my new machine, and then I > > ran the commands as follows(ran as root): > > > > yum install qemu-kvm > > yum install qemu > > yum install libvirt > > yum install virt-manager > > > > then I started the libvirtd as follow: > > /etc/init.d/lbvirtd start > > > > So , it ran well. > > > > > > However, after a few days , when I ran the command "virt-manager" , it > > showed as bellow: > > > > > > > > it seemed that ti cannot connect the hipervisor. > > This list is plain-text only. Your embedded graphic (at least, I'm > assuming you attempted to embed a graphic) got eaten by the list server, > so we don't know what message you got. Can you paste the actual error > message you got? > > Meanwhile, we have a wiki page for diagnosing common problems. Perhaps > your issue is covered by one of the items there; and if not, it would > help if you could add a new page with the steps you take in resolving > the issue. http://wiki.libvirt.org/page/Troubleshooting > > > > > > > Then I did a little test: > > I ran the conmand "virsh list" , but it didnot work,either. > > Again, knowing the exact output of 'virsh -c qemu:///system list', > rather than just saying it didn't work, will help others in diagnosing > your issue. > > -- > Eric Blake eblake(a)redhat.com +1-919-301-3266 > Libvirt virtualization library http://libvirt.org > >

1 0

[libvirt-users] Unable to connect qemu:///system on Fedora16 X86_64
by suyi wang 28 Feb '12

28 Feb '12

Dear all: I ust installed Fedora16Fedora16 X86_64 on my new machine, and then I ran the commands as follows(ran as root): yum install qemu-kvm yum install qemu yum install libvirt yum install virt-manager then I started the libvirtd as follow: /etc/init.d/lbvirtd start So , it ran well. However, after a few days , when I ran the command "virt-manager" , it showed as bellow: it seemed that ti cannot connect the hipervisor. Then I did a little test: I ran the conmand "virsh list" , but it didnot work,either. Then I used Python as follows: import libvirt cc = libvirt.open(r'qemu:///system') but it didnot work,either. All the test above are under root . I hope that someone can give me some suggestiion. Thanks to all! Suyi Wang

2 1

[libvirt-users] Bug? virGetLastError() returns NULL after virDomainCreate() if Domain running already
by Guido Winkelmann 28 Feb '12

28 Feb '12

Hi, As the subject line indicates, when I'm calling virDomainCreate() to start a defined domain, and that command fails because the domain is already running at that point, virGetLastError() will return NULL, instead of a proper pointer to an error code. Libvirt will, however, print an error to the console: libvir: QEMU error : Requested operation is not valid: domain is already running This looks like a bug to me, and I'm fairly certain that earlier version would return an error code of VIR_ERR_OPERATION_INVALID at that point. Or maybe that was just virDomainDestroy(), and I'm just blindly assuming that virDomainCreate() will work analogously... Anyway, no error code is bad, because I can't tell what happened. This is version 0.9.8. I couldn't find anything about this issue in the changelogs for 0.9.9 and 0.9.10, so I'm assuming the issue has not been addressed since. Guido

2 6

[libvirt-users] Sys::Virt
by Evaggelos Balaskas 28 Feb '12

28 Feb '12

Can anyone help me: $ perl Makefile.PL Checking if your kit is complete... Looks good Writing Makefile for Sys::Virt Writing MYMETA.yml and MYMETA.json $ make /usr/bin/perl "-Iblib/arch" "-Iblib/lib" perl-Sys-Virt.spec.PL perl-Sys-Virt.spec make: *** [perl-Sys-Virt.spec] Segmentation fault make: *** Deleting file `perl-Sys-Virt.spec' Evaggelos Balaskas - Unix System Engineer http://gr.linkedin.com/in/evaggelosbalaskas

2 5

[libvirt-users] Not able to mount iSCSI volume to Guest VM
by Sarath P R 25 Feb '12

25 Feb '12

Hi All, I have created a 10GB logical volume in iSCSCI target machine. I am abe to mount the volume in my Dom0 . But I am not able to attach the volume to Guest VM I tried "<pool type='iscsi'>"+ "<name>virtvolume</name>"+ "<source>"+ "<host name='192.168.1.1'/>"+ "<device path='"+src+"'/>"+ "</source>"+ "<target>"+ "<path>/dev/disk/by-path</path>"+ "</target>"+ "</pool>" I Could create a storage pool using the above xml. But how can I attach that volume to the guest VM. Please Help. Thanks in Advance -- Thank You Sarath P R

1 0

[libvirt-users] lockmanager for use with clvm
by Michael Menge 25 Feb '12

25 Feb '12

Hi, i am setting up a cluster of kvm hypervisors managed with libvirt. The storage pool is on iscsi with clvm. To prevent that a vm is started on more than one hypervisor, I want to use a lockmanager with libvirt. I could only find sanlock as lockmanager, but AFSIK sanlock will not work in my setup as I don't have a shared filesystem. I have dlm running for clvm. Are there lockmanager implementations that use dlm? Or other implementations that can be used in my setup? Regards Michael Menge -------------------------------------------------------------------------------- M.Menge Tel.: (49) 7071/29-70316 Universität Tübingen Fax.: (49) 7071/29-5912 Zentrum für Datenverarbeitung mail: michael.menge(a)zdv.uni-tuebingen.de Wächterstraße 76 72074 Tübingen

3 2

[libvirt-users] Regarding file vs block devices for the guest
by Udayan Bapat 24 Feb '12

24 Feb '12

Hi There is a general consensus in regards to use of a block device (exported partition) over a file device (.img or .qcow2) image. It is expected to see a big performance jump in case of a block device. However in my test I have seen it is not always the case. My testbed has Fedora-14.0 as the host and Ubuntu-10.0 as guest running on Intel x86 4 core/8 thread VTX enabled system. I ran file system benchmarks from the guest using iozone and measure throughput as well as CPU utilization. I observed that upto device sizes of 8 GB, file device (.img with ext3 partition) consistently outperformed block device (ext3 partition exported directly to the guest). Is similar anomaly seen by the community too? If not, i.e. block device should always outperform file, which layer should I be looking at to understand what could be happening. Just to make sure, I repeated same test a bunch of times to eliminate noise. As I increased the HDD size, then block device performance drastically improved and consistently performed better than fileIO. In both tests, my guest had 512 MB RAM with 1 vcpu. I ran all 11 iozone test to do performance comparisons. Any pointer/insight would be extremely helpful. Thank you in advance. -- Udayan

1 0

Re: [libvirt-users] [SeaBIOS] RFE: Print amount of RAM
by Igor Galić 23 Feb '12

23 Feb '12

Cross posting from seabios@ [snip] > > Now I'd like to ask you guys if you could print the amount > > of RAM available -- like most other BIOSes out there already > > do - it would be extremely helpful to aid troubleshooting. > > SeaBIOS boots too fast for anything useful to be seen on the screen. > That said, one can add "-chardev stdio,id=seabios -device > isa-debugcon,iobase=0x402,chardev=seabios" to the qemu command line > to > see the SeaBIOS debugging output. That output has the memory info > and much more. I guess that's then something to propose for libvirt as RFE as an extra argument for <bios useserial='yes' /> we could introduce <bios useserial='debug' /> perhaps. > -Kevin Thanks Kevin. So long, i -- Igor Galić Tel: +43 (0) 664 886 22 883 Mail: i.galic(a)brainsware.org URL: http://brainsware.org/ GPG: 6880 4155 74BD FD7C B515 2EA5 4B1D 9E08 A097 C9AE

1 0