- Users - libvirt List Archives

[libvirt-users] Seg fault in 0.9.4-rc1
by Whit Blauvelt 28 Jul '11

28 Jul '11

Hi, Using virt-manager, asking to migrate a kvm domain from the same box virt-manager is running on to another consistently results in libvirtd faulting on the local box. Migration in the other direction, from the second box to the box virt-manager is on, works fine. A libvirtd log made with debug=1 is at http://transpect.com/debug/libvirtd_segfault.log It was doing the same thing with a recent 0.9.3 git version too. This is with virt-manager 0.9.0. Whit

2 2

[libvirt-users] qemu:///session vs qemu:///system
by Peter Thomassen 28 Jul '11

28 Jul '11

Hi, I am running several QEMU VMs on a host machine. There's different persons being responsible for different VMs. I would like to find a way to allow everybody to manage just the VMs they are assigned to (i.e. insert some ISO file, stop the machine and so on) without allowing them to disturb the other VMs. Also, I would like to allow these persons to set up new VMs (there's plenty of unallocated storage) without having to ask me (root) to do it for them, but without allowing them to delete other VMs. (Just setting user rights on the filesystem is not enough, as I use LVM volumes which are not assigned to a user in the typical sense, are they?) I know that most of this could probably be achieved by using qemu:///session instead of qemu:///system. However, I read that there's networking issue in this case (which issues precisely?), and I am also not sure if there's other disadvantages like less efficient handling in some situations by the kernel, or whatever. Is anybody there who could clarify this for me? Thank you, Peter

1 0

[libvirt-users] small webGUI?
by Stefan G. Weichinger 28 Jul '11

28 Jul '11

Greetings, at a customer we run a single VM for a specific application. Unfortunately this VM does not start automatically with the host although I checked the relevant box within virtual machine manager. I assume upgrading libvirt might help (we run 0.8.7-r1 right now, gentoo linux). Aside from that it would be good to have a simple WebGUI for the users to let them press START just in case. They run Windows on their machines so I can't give them VMM ... Could you point me at the most simple GUI to do that? I'd prefer something very lightweight without too much requirements (for example I'd like to avoid installing mysql for that ;-) ). Thank you, Stefan

1 2

[libvirt-users] with current libvirt git virsh expects vbox, refuses kvm define
by Whit Blauvelt 28 Jul '11

28 Jul '11

Okay, just installed libvirt from today's git on two Ubuntu systems. The first one went fine (except for continuation of the problem that virsh and virt-manager cannot connect to 0.8.3 systems that's been there in the 0.9.x libvirts) ... but on the second, after install, "virsh list --all" presents nothing. And: # virsh define tu1004.xml error: Failed to define domain from tu1004.xml error: internal error unexpected domain type kvm, expecting vbox These are _exactly_ the same xml files as are still working on the first system. And in both cases the new libvirt was built with defaults. Long ago the second system did have VirtualBox on it, so that may have been found while it was setting up config, but that's no good reason for KVM to fail here, right? Must be a thing with the git. Ran the 0.9.3 release built on this system previously without this problem. Whit

5 11

Re: [libvirt-users] VirtualBox headless
by Guilherme Santos 27 Jul '11

27 Jul '11

My objective is also to run headless virtual machines through libvirt library. I am using an Ubuntu 10.04 distro and libvirt 0.9.2. Before asking you guys about it, I searched on the web and found this: Hi, > > thanks for your hint. > > I read the lines in the documentation, but unfortunately editing the xml > > within virsh lead to the errors. > > The Solution to a headless domain was to edit the XML file from bash and > then > > #virsh: define <domainname.xml> > > to reload the new values for the graphics type of the machine. > > Obvioulsy something went wrong when editing "live".... > > Regards, > Christian > > Am 24.06.2010 18:27, schrieb Jean-Baptiste Rouault: > > Hi, > > > > Something like this should work, as can be seen in the > > > documentation here : http://libvirt.org/formatdomain.html#elementsGraphics > > > > <drivers> > > <graphics type='rdp' autoport='yes' /> > > > </drivers> > > > > Regards, > > > > Jean-Baptiste > > > > > >> Hello, > >> > >> I'm trying to use the VirtualBox/vbox driver together with libvirt 0.8.1 > >> on a Ubuntu Lucid Machine, VirtualBox Release 3.1. > > >> > >> The goal is to run the virtual machines headless and to connect using RDP. > >> > >> When starting a virtual domain using 'virsh start <domainname>' the GUI > >> of VirtualBox will pop up and I see no option to prevent this. If X is > > >> not running, the virsh command will terminate with 'unknown failure'. > >> > >> I've found some hints on inclusion of the headless statement in the > >> 'graphics' part of the XML definition file, but when I tree to adopt > > >> this, the editing of the file will crash. > >> > >> Has anyone achieved to manage VBox machines with libvirt on a headless > >> server ? > >> > >> Thanks in advance for some hints or experiences with this issue, > > >> Christian > >> > > Unfortunately this isn't working for me. Following is my XML configuration: <domain type='vbox'> <name>THEDOMAIN</name> <uuid>33f6e188-1479-46f4-9d90-4d5a2d3a30f2</uuid> <memory>393216</memory> <currentMemory>393216</currentMemory> <vcpu>1</vcpu> <os> <type arch='i686'>hvm</type> <boot dev='hd'/> <boot dev='cdrom'/> </os> <features> <acpi/> </features> <clock offset='localtime'/> <on_poweroff>destroy</on_poweroff> <on_reboot>destroy</on_reboot> <on_crash>destroy</on_crash> <devices> <disk type='file' device='disk'> <source file='/local/vbox-shared.vdi' /> <target dev='hda' bus='ide'/> </disk> <disk type='file' device='cdrom'> <source file='/local/ubuntu-11.04-desktop-i386.iso' /> <target dev='hdc' bus='ide'/> </disk> <interface type='user'> <mac address='08:00:27:9a:d0:63'/> <model type='Am79C973'/> </interface> <input type='mouse' bus='ps2'/> <graphics type='rdp' autoport='yes' /> <video> <model type='vbox' vram='12' heads='1'> <acceleration accel3d='no' accel2d='no'/> </model> </video> </devices> </domain> When I run command *virsh start THEDOMAIN* it is popping up the following error message: *OpenRemoteSession/LaunchVMProcess failed, domain can't be started* As it wasn't working I also tried other options, but unfortunately, using all other graphic type options (except for 'spice', which I didn't try) the GUI is launched. I don't think I'll be needing a RDP connection, neither anything similar to that, except if these are necessary to basic management of the vm through virsh (start, stop, reboot, create and revert snapshots). Isn't there one simple way to run a virtual machine domain headless without vnc, spice or rdp servers? If not, which could be going wrong in my xml configuration? To help understanding of my desired behavior, what I need is to do through virsh what can be done using *VBoxManage startvm THEDOMAIN --headless* or *VBoxHeadless startvm THEDOMAIN*, because will use other hypervisors in the future, so it is a critical thing to me. Thanks in advance for the help, Guilherme Santos

2 1

[libvirt-users] Wanted: method for qemu hook script to know if called for migration
by Whit Blauvelt 27 Jul '11

27 Jul '11

Hi, "Before a QEMU guest is started, the qemu hook script is called" (http://libvirt.org/hooks.html) In writing a simple bash script to prevent a VM on shared storage from being started on one host if it's already running on another, it's easy enough to check if virsh on the other host reports the particular VM as "shut off" there, and go ahead if that's the case. What's I'm having trouble coming up with is a simple method to allow a migration from the second system over to the first with that script in place. As the doc says, /etc/libvirt/hooks/qemu is "Executed when a QEMU guest is started, stopped, or migrated," so when it's simply being started, it makes sense to be sure it's not already running on the second system, and prevent the start if it's not. Works fine for me. But when a migration is requested, the script is called in the same way as when it's being started. What I need is a way, from within the script, to tell whether it's a migration context, so that then the start will be allowed, despite the VM's already being running on the second system. Any ideas how to do that? Unfortunately what's passed to the script directly is just "[vmname] start begin -" in both cases. And I don't see that a separate migrate process shows up in the process tables until after getting beyond the qemu hook script. As a feature suggestion, it would be nice if that fourth input variable, currently "-", was something like "migrate" in this instance. Thanks, Whit

3 5

[libvirt-users] python-libvirt for 0.9.3 leaking file descriptors
by Igor Serebryany 26 Jul '11

26 Jul '11

i've reported this issue before, so i guess this is a regression. looks like the python bindings for 0.9.3 are leaking file descriptors: root@cloud1:~# python Python 2.6.6 (r266:84292, Dec 26 2010, 22:31:48) [GCC 4.4.5] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> import libvirt >>> libvirt.getVersion() 9003 >>> def openclose(): ... c = libvirt.open('qemu:///system') ... c.close() ... >>> import os >>> os.system('lsof -p %d | wc -l' % os.getpid()) 37 >>> openclose() >>> os.system('lsof -p %d | wc -l' % os.getpid()) 40 >>> openclose() >>> os.system('lsof -p %d | wc -l' % os.getpid()) 43 every time I open and close a connection to libvirt, i loose 3 file descriptors. these are a socket and two fifo pipes. i'm using the -3 build on debian sid. can someone test this on a fedora machine? curious if this is something inside libvirt or a problem with the build... thanks, --igor

4 3

[libvirt-users] insert ISO into a KVM guest
by Jeff 26 Jul '11

26 Jul '11

There's got to be a simple way to attach an ISO to a KVM guest. I just haven't yet discovered it. I've tried adding the following stanza to my guest XML file, under the <devices> stanza and starting the guest: <disk type='file' device='cdrom'> <driver name='qemu' type='raw' /> <source file='/etc/libvirt/qemu/isos/kvm1.iso' /> <target dev='hda' bus='ide' /> <readonly /> </disk> But I get the following: error: Failed to start domain rhel6guest error: internal error process exited while connecting to monitor: char device redirected to /dev/pts/3 qemu: could not open disk image /etc/libvirt/qemu/isos/kvm1.iso: Permission denied A perms issue should be easy to address one would think, but I've tried every possible combination of perms changes including 777. Changed ownership from root to qemu:qemu (though I'm running as root). I'm not running SELinux. Anyone have any insights on how to mount a simple ISO into a KVM guest? Both at build time AND while the guest is running? Thanks Jeff

2 1

[libvirt-users] Ping not working after a migration
by Adnan Pasic 26 Jul '11

26 Jul '11

Hello guys, maybe you can help me with following issue. I have created a little cloud with a host and two worker nodes using opennebula. The setup went successfully until now, I am able to create VM's and move them via normal and live migration. Another (possibly) important information is that I configured my virtual bridge on both worker nodes like this: auto br0 iface br0 inet static address 192.168.0.[2|3] netmask 255.255.255.0 network 192.168.0.0 broadcast 192.168.0.255 #gateway 192.168.0.1 bridge_ports eth0 bridge_stp on bridge_maxwait 0 The command "brctl show" gives me following things back: bridge name bridge id STP enabled interfaces br0 8000.003005c34278 yes eth0 vnet0 (<- only appears on node with running VM) virbr0 8000.000000000000 yes According to thelibvirt wiki this setting is good as is. However, the issue I'm having is that when I create a VM and assign a static IP to it, which looks like e.g. 192.168.0.5, I firstly am able to ping this VM from both worker nodes, and also when I perform a live migration the ping stops for a few seconds (until the nodes realize the new route to this VM) and then starts pinging normally again. However, when I perform a normal migration the ping doesn't recover anymore, but answers repeatedly with: Destination Host Unreachable Do you know what could be the problem? Where is the difference between a normal and live migration and how can the ping after live migrating still work, but after a normal migration not? Thanks a lot! Regards, Adnan

1 1

[libvirt-users] Split a qcow2-image?
by Stefan G. Weichinger 26 Jul '11

26 Jul '11

Hello, I converted a W2003-Server to a vmdk-image, I didn't think too much back then ... Now I have 2 windows-drives C: and E: inside one vmdk-image-file, which was converted to qcow2. I have installed drivers and stuff into that VM already and would like to avoid to start over again with the vmdk-file -> Is there a way to split that one qcow2 into 2 separate files, one for each windows-harddisk? or plan B: can I possibly enlarge the part of C: inside that qcow2? I use qemu-kvm on gentoo linux, and libvirt-0.9.3-r1. I would appreciate any help on this topic! Thanks Stefan

2 3