- Users - libvirt List Archives

[libvirt-users] VirtualBox headless
by Christian Patsch 29 Jun '10

29 Jun '10

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

2 2

[libvirt-users] can't access VM after Fedora 13 KVM install
by Joel Uckelman 28 Jun '10

28 Jun '10

Hi, I'm installing a Fedora 13 virtual host on a Fedora 13 system, like this: virt-install --ram 2048 --vcpus 4 --name test --os-type linux \ --os-variant fedora13 --disk /var/lib/libvirt/images/test.img,size=10 \ --cdrom /var/lib/libvirt/Fedora-13-x86_64-DVD.iso --vnc --noautoconsole I connect to the installer using virt-manager, and the install appears to complete successfully. Then, I reboot the virtual machine and reconnect with virt-manager. What I see is this: Starting SeaBios (version 0.5.1) gPXE (http://etherboot.org) - 00:04.0 C900 PCI2.10 PnP BBS PMM7FE0@10 C900 Booting from Hard Disk... If I connect to the VM's console using 'virsh console', there's no output at all. I don't see anything in /var/log/messages or /var/log/libvirt that looks like an error. So, I'm not sure what's going on here, as this is the first time I've done anything with virtualization. Has my VM hung, or is this what's supposed to happen? I was expecting that somewhere along the line that my VM would get an IP address via DHCP so I could SSH to it, but don't see anything indicating that has happend, or what the IP address is if it has. Is something broken here, or have I missed a crucial step? -- J.

1 1

[libvirt-users] network interface management in bridge firewall configuration
by Aleksander Trofimowicz 28 Jun '10

28 Jun '10

Hello, I'm just wondering why I can't manage my network interfaces through libvirt when the following kernel parameters are turned on: net.bridge.bridge-nf-call-ip6tables net.bridge.bridge-nf-call-iptables net.bridge.bridge-nf-call-arptables Is it a bug or by design? If the latter, could someone explain me premises of such decision? I'm aware of security implications of mixing conntrack and bridge bits, so we can skip that point. This behaviour is noticeable when using: libvirt-0.8.1-1.fc13.x86_64 netcf-libs-0.1.6-1.fc13.x86_64 -- thanks, aleksander trofimowicz

2 1

[libvirt-users] sound problem
by Stephen Liu 25 Jun '10

25 Jun '10

Hi folks, KVM - 1:0.12.4+dfsg-1~bpo50+1 0 host - Debian 5.0 I have been suffering on sound problem for sometimes. Please advise whether the same has been solved. If YES please advice how to solve it. TIA B.R. Stephen L

1 0

[libvirt-users] i need help
by Jounewe Lambang Paul Roland 24 Jun '10

24 Jun '10

I am french student who works on a libvirt projet, it consists to get information from an hypervisor by using libvirt API in perl in cgi script but i have a great problem. I wrote my script and its works in EPIC but when i am trying to run it from apache web server as a CGI script, it goes wrong and i have this message: libvirt code 38 message: can no recv data: ssh connect to host myZenhost.mydomaine.fr port 22 in my browser. i made libvirtd installation and tried command line virsh -c xen+ssh://root@myZenhost.mydomain.fr wich works correctly i dont know what wrong.I run the script to remote xen hypervisor as root with a good config of ssh (transport mode), I realize that the environment variables of root account (client) is not the same with apache webserver, and i am set the same path in apache env variable but there are no change. please i need your help to fix this problem. Cdt. Paul Roland JOUNEWE LAMBANGIngénieur Réseaux TélécomMob:+336 03 74 60 565, Av Albert 1er 95600 Eaubonne-France le mensonge est comme une bulle d'air qu'on tente désespérément de retenir au fond d'une récipient plein d'eau, tôt ou tard il finit par remonter en surface.

1 0

[libvirt-users] Domain did not show up error
by Matt Wallace 24 Jun '10

24 Jun '10

Hi all, I've run into an issue with using libvirt via the ruby bindings. When trying to create a new domain via xml, I get the following error: libvir: QEMU error : internal error Domain webserver.staging.myco.com didn't show up When I check using virsh, the domain has been defined and shows as "running" when I execute "list --all" but does not show up when I execute "list" nor will it let me connect via "console webserver.staging.myco.com". Can anyone shed any light on why this might be? O/S: Centos 5.4 Ruby: ruby 1.8.5 (2006-08-25) [x86_64-linux] Libvirt: 0.6.3 If you need any more information, please let me know. If you want to see the code that is being used, it's cucumber-vhost (http://github.com/proffalken/cucumber-vhost) Thanks, Matt

1 0

[libvirt-users] Obtaining UTC Time from Remote VM Host
by Development 24 Jun '10

24 Jun '10

Quick question on whether it is possible to use the libvirt API exposed by the libvirt DLL compiled on a Vista machine to obtain the system time in UTC of the remote machine hosting several VMs? Many thanks...

2 2

[libvirt-users] Qemu backend: Cannot start new domains after some time
by Guido Winkelmann 24 Jun '10

24 Jun '10

Hi, After using libvirt+qemu for a while, I'm finding that I can no longer reliably start new domains. Most of the time when I try, I'm getting this, or something like it: virsh # start dhcpv6server error: Failed to start domain dhcpv6server error: monitor socket did not show up.: No such file or directory The same thing happens when trying to start the domain using the C-API. This happens particularly often when there are any domains running that are using a lot of CPU, even when most of the physical cores of the host machine are still sitting idle. (I've created a number of Linux VMs, with one CPU core each, that do nothing but calculate primes in an endless loop for load-testing purposes.) If I destroy all of the high-load domains, I can usually start up new domains again. The time between requesting the domain startup and the error message being printed varies greatly, from less than a second to more than a minute. The host machine is running Fedora 12 with libvirt 0.8.1 and qemu-kvm 0.12.4. Any ideas what might be causing this, and what can be done about it? It would be a big problem for us if we cannot start any new domains just because some of the existing ones are using a lot of CPU... Regards, Guido

1 0

Re: [libvirt-users] Question about the remote Authentication in the libvirt Java bindings
by Bryan Kearney 22 Jun '10

22 Jun '10

Dustin: I am adding this response to the libvirt list as well so that others may see it. The authorization is done via a callback. The ConnectAuthDefault() that you are using is one such implementation. You could create your own subclass of ConnectAuth, and override the method: public abstract int callback(Credential[] cred); And then provide the username/password in that fashion. Look at the ConnectAuthDefault in the case statement how it chooses to ask for username and password. -- bk On 06/21/2010 10:35 PM, Dustin Xiong wrote: > Dear Bryan Kearney： > > I hava a question about the remote authentication of the libvirt java > bindings. I run the test.java of the libvirt-java-0.4.3.tar.gz. I can > connect the remote host use the > > ConnectAuth defaultAuth = new ConnectAuthDefault(); > try{ conn = new Connect("xen+ssh://192.168.0.146/", defaultAuth, 0); > > Then I input the remote host's ssh password, I can connect the host. > If I don't want to input the ssh password, I want write the password > into the libvirt or libvirt-java' src code. > How can modify the libvirt or libvirt-java src code, the libvirt will > auto connect to the remote host without user input the ssh code. > > Thanks for help. > > -Dustin

1 0

[libvirt-users] How to get supported feature list in driver
by Ryota Ozaki 21 Jun '10

21 Jun '10

Hello Is there an easy way to get supported feature (API) list in each driver? The function would be to return a list which describes which APIs are supported and which APIs are not. Thanks in advance ozaki-r

2 2

[libvirt-users] libvirtError: poll on socket failed: Interrupted system call
by jeremy avnet 19 Jun '10

19 Jun '10

Hello. I just upgraded from libvirt 0.7.0 to 0.7.5 in Ubuntu Lucid and now my application will get into a hung state. I have to kill it at which point I get the following error from the Python bindings: Traceback (most recent call last): File "/usr/local/app/vm.py", line 246, in launch vm.create() File "/usr/lib/python2.6/dist-packages/libvirt.py", line 300, in create if ret == -1: raise libvirtError ('virDomainCreate() failed', dom=self) libvirtError: poll on socket failed: Interrupted system call Some searches turned up a patch involving signals sent by ncurses apps resizing, but this is not an ncurses app. It's a daemonized Python Twisted application. Unfortunately, this problem is making it so the application hangs indefinitely until we kill it and start it again. Any suggestions as to what might be causing this? Thanks! .:. jeremy

1 0

[libvirt-users] java.lang.UnsatisfiedLinkError: jnidispatch (/com/sun/jna/linux-i386/libjnidispatch.so) not found in resource path
by Rafael Braga 18 Jun '10

18 Jun '10

Hi everybody, I'm trying to run the sample of the site, but is not working. In fact, already includes jna.jar in my class path, see. What can it be? -- Rafael Braga

1 1

[libvirt-users] Snapshots - can't take a snapshot with running VM
by Julian Robbins 17 Jun '10

17 Jun '10

Hi I'm quite new to KVM/libvirt0 so hope you can help I'm using Ubuntu Lucid 10.04 and libvirt0 0.8.1 When I try a :- snapshot create vm_name I get an error, ie 'error -5 while writing' Googling doesnt bring anything up. Taking snapshots of a vm when it isnt running works fine. Can I really only create snapshots of a vm when it isnt running ? -- Kind regards Julian

2 4

[libvirt-users] cannot acquire state change lock
by Julian Robbins 17 Jun '10

17 Jun '10

Hi I keep getting libvirtError: Timed out during operation: cannot acquire state change lock errors - what does this mean exactly ? Using Ubuntu 64 bit 10.04 host and guest -- Kind regards Julian Robbins

2 1

[libvirt-users] libvirt API hangs
by Zhang Qian 17 Jun '10

17 Jun '10

Hi, I have someVMs in my KVM hosts(RHEL 5.5), I found when run "virsh save <domain> <file>" to save a VM, the command "virsh dominfo <domain>" will hang until the save command finishes. Is this libvirt's behavior? Or KVM/QEMU's? This may impact the performance. Thanks, Qian

2 1

[libvirt-users] Bridge with macvtap problem
by Eduardo 16 Jun '10

16 Jun '10

Hi friends! I'm using libvirt 0.8.1 with qemu-kvm 0.12.4 on Linux 2.6.34. At the moment there are 2 virtual machines running inside that host without problem. They are using the macvtap feature for networking share. But I can't do ping from the host to the guest, but I need this. And whatever communication between them. Ping with other host is normal. Below follows the xml domain definition of one guest: <domain type='kvm' id='2'> <name>vox2</name> <uuid>2e143f6e-5506-8073-c5fc-3ea92fc2b579</uuid> <memory>524288</memory> <currentMemory>524288</currentMemory> <vcpu>2</vcpu> <os> <type arch='x86_64' machine='pc-0.12'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> </features> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <emulator>/usr/bin/qemu-system-x86_64</emulator> <disk type='file' device='disk'> <source file='/virtual/vox2.qcow2'/> <target dev='sda' bus='scsi'/> <alias name='scsi0-0-0'/> <address type='drive' controller='0' bus='0' unit='0'/> </disk> <controller type='scsi' index='0'> <alias name='scsi0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </controller> <interface type='direct'> <mac address='52:54:00:2f:fe:3d'/> <source dev='eth0' mode='bridge'/> <target dev='macvtap1'/> <model type='virtio'/> <alias name='net0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </interface> <input type='mouse' bus='ps2'/> <graphics type='vnc' port='5901' autoport='no' keymap='pt-br'/> <video> <model type='vmvga' vram='9216' heads='1'/> <alias name='video0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> </devices> </domain> P.S.: The default network of libvirt is disabled. -- *Eduardo Ramos* Asterisk Manager Linux User www.freedominterface.org +55 12 81129581 exten=>eduardo,1,hangup()

1 0

[libvirt-users] howto search this mailing list archive?
by Julian Robbins 16 Jun '10

16 Jun '10

Another quickie I've long and hard (honest !), but can't find a way to search this ml's archives. How do you do this ? -- Kind regards Julian Robbins

2 1

[libvirt-users] apparmor and kvm/libvirt
by Julian Robbins 16 Jun '10

16 Jun '10

Hi After kindly answering my question on snapshots not working whilst vm is on, i find that apparmor is still disturbing the process where snapshots can get written But although i should just be able to disable apparmor for my vm in virt-manager, whilst vm is off; that when i start the vm the apparmor settings go back to being enabled and set to dynamic This is stopping me from taking snapshots again whilst vm is on. :-( I have also tried disabling the apparmor profile for libvirt on the host but this appears to have had no effect. I know very little about appamor and it is not particularly important to be used in this case. I am using Ubuntu 10.04 LTS on both host and guest 64 bit. -- Kind regards Julian Robbins

1 0

[libvirt-users] SCM_RIGHTS file error when attaching device?
by steven765＠yahoo.com 16 Jun '10

16 Jun '10

So I just upgraded to 0.8.1-1 and virt-manager 8.4 when i try to attach my network card that's been perfectly fine in the past i now get... virsh # attach-device gnome /etc/libvirt/qemu/nic.xml error: Failed to attach device from /etc/libvirt/qemu/nic.xml error: internal error unable to send TAP file handle: No file descriptor supplied via SCM_RIGHTS Any thoughts?

2 1

[libvirt-users] How to configure kvm development environment
by Avdhoot Joshi 15 Jun '10

15 Jun '10

Hi, Installed RHEL 5.5 on AMD Opteron. [root@wlng-ppcore-nd3 ~]# uname -a Linux wlng-ppcore-nd3 2.6.18-194.el5 #1 SMP Tue Mar 16 21:52:39 EDT 2010 x86_64 x86_64 x86_64 GNU/Linux [root@wlng-ppcore-nd3 ~]# virsh version Compiled against library: libvir 0.6.3 Using library: libvir 0.6.3 Using API: QEMU 0.6.3 Running hypervisor: QEMU 0.9.1 This is the default version of virsh using libvirt 0.6.3. Downloaded libvirt 0.8.1, configured it as follows: $ configure $ make $ make install What should be done so that virsh uses the latest 0.8.1 libvirtd binary, how to completely stop 0.6.3? I want to execute a program using virDomainMemoryStats which is available only in latest libvirtd. Thanks, Avdhoot

1 0

Re: [libvirt-users] Problems moving qemu machine to libvirt
by Gunnar Thielebein 15 Jun '10

15 Jun '10

Hi Erik, thanks for the solution provided. I tried that but unfortunately it spits out this error message: > Traceback (most recent call last): > File "/usr/share/virt-manager/virtManager/engine.py", line 588, in run_domain > vm.startup() > File "/usr/share/virt-manager/virtManager/domain.py", line 150, in startup > self._backend.create() > File "/usr/lib/python2.6/dist-packages/libvirt.py", line 300, in create > if ret == -1: raise libvirtError ('virDomainCreate() failed', dom=self) > libvirtError: internal error cannot parse QEMU version number in '' Looks to me that libvirt does some version checks with the emulator-node and don't get reliable output. The wrapper script: /etc/libvirt/kvmvde: > #!/bin/bash > exec /usr/bin/kvm "$@" \ > -net nic,vlan=0,model=virtio,macaddr=52:54:61:44:21:23 \ > -net vde,vlan=0,group=libvirtd,mode=0770,sock=/var/run/vde.ctl The xml section: ... > <devices> > <emulator>/etc/libvirt/kvmvde</emulator> > <disk type='file' device='disk'> > <source file='/var/lib/libvirt/images/vlxp01-gt/tmp7Jiyzz.qcow2'/> > <target dev='hda' bus='ide'/> > </disk> ... Btw. is the bridge entry really necessary in the xml file? I wonder if this wouldn't also work, but I am not at this step, currently: => http://libvirt.org/formatdomain.html#elementsNICSSlirp "Generic ethernet connection" > ... > <devices> > <interface type='ethernet'/> > ... > <interface type='ethernet'> > <target dev='vnet7'/> > <script path='/etc/qemu-ifup-mynet'/> > </interface> > </devices> > ... Perhaps this is a place to also start the vde hub? On 10.06.2010 13:34, Erik de Castro Lopo wrote: > #!/bin/bash > exec /usr/bin/kvm $@ \ > -net nic,vlan=0,model=rtl8139,macaddr=52:54:61:44:21:23 \ > -net vde,vlan=0,group=vde2-net,mode=0770,sock=/var/run/kvm-vde-ctl Any reason no to use the virtio driver? Best, Gunnar

1 0

[libvirt-users] Functionality of virsh net-edit
by Arseny Klimovsky 15 Jun '10

15 Jun '10

Hi, I am using version libvirt on debian lenny from backports repository, version 0.7.6-1~bpo50+1, amd64. Th functionality of "virsh net-edit" command is not very clearly documented, and I think it should change a configuration of a running network. But I see it doesn't. So, I want to add new host, for example: #I have old configuration root@xxx:/etc/libvirt/qemu/networks# virsh net-edit default Network default XML configuration edited. #Here I added new line: new host definition root@xxx:/etc/libvirt/qemu/networks# cat default.xml #Here I see that new line root@xxx:/etc/libvirt/qemu/networks# virsh net-dumpxml default root@xxx:/etc/libvirt/qemu/networks# virsh net-edit default #Here I see exactly the old network definition Is this behaviour expected? If yes, how to just add new host on a running network? -- Arseny Klimovsky

2 1

[libvirt-users] Documentation Bug: struct virSchedParameter missing value field
by Guido Winkelmann 14 Jun '10

14 Jun '10

Hi, The libvirt documentation under http://libvirt.org/html/libvirt-libvirt.html#virSchedParameter fails to list the "value" member, even though that is kind of essential for using the struct. This is what the documentation says: struct virSchedParameter{ charfield[VIR_DOMAIN_SCHED_FIELD_LENGTH] field : parameter name int type : parameter type } This is how it looks in the actual header file: typedef struct _virSchedParameter virSchedParameter; struct _virSchedParameter { char field[VIR_DOMAIN_SCHED_FIELD_LENGTH]; /* parameter name */ int type; /* parameter type */ union { int i; /* data for integer case */ unsigned int ui; /* data for unsigned integer case */ long long int l; /* data for long long integer case */ unsigned long long int ul; /* data for unsigned long long integer case */ double d; /* data for double case */ char b; /* data for char case */ } value; /* parameter value */ }; This one had me scratching my head for a while, because the way it is documented, it's absolutely not clear how the struct is supposed to be used. Only after studying the virsh source code for a while did I notice that this struct apparently has more members than the documentation lets on... Guido

1 0

[libvirt-users] Missing target dev for interface
by jeremy avnet 11 Jun '10

11 Jun '10

I have over 20 guests running on a particular host and the three longest running ones are missing the network device information in their XML description: <interface type='network'> <mac address='54:52:10::f2:c5'/> <source network='local'/> </interface> Compared to the others which have: <interface type='network'> <mac address='54:52:d1:32:0a:a1'/> <source network='local'/> <target dev='vnet13'/> </interface> This is problematic as I wanted to use domifstat on the domains. Using the packet counts shown via ifconfig on the host, I was able to determine what the vnet interface should be, but passing that to domifstat doesn't work: $ virsh domifstat maki221 vnet13 Connecting to uri: qemu:///system error: Failed to get interface stats guest19 vnet13 error: invalid argument in invalid path, 'vnet13' is not a known interface How did this happen? Is there a way I can get libvirt to know about the interfaces? And the guest OS's network connections work fine. Thanks! .:. jeremy

2 2

[libvirt-users] Multipath pools - practical for use?
by Justin Clift 10 Jun '10

10 Jun '10

Hi all, Writing up some Fedora documentation, and looking to figure out the best way of mapping multipath network(!) LUNs to pools in libvirt. ie Infiniband SRP LUNs, but would probably apply to Fibre Channel equally as well. There are two approaches I can think of easily: a) Large LUNs (ie TB+) that are mapped to a host server as disk, with each LUN being configured as an LVM pool. This lets the host slice the pool up into volumes with the libvirt volume creation commands. b) Pre-purposed LUNs mapped to a host server as disk, but passed through directly to their respective guests. ie Something like Oracle, with 3 SAN LUNs mapped to guest FOO for its database data In a non-libvirt world I'd look at configuring the LUNs on the host with persistent name using multipath, then present them appropriately to whatever needs them from there. However, libvirt has a pool type called "multipath". Looking at the docs online, it doesn't appear to be all that fully featured at this point. Seems more of a brief mention and footnote, so I'm wondering whether to use it or skip it for now in the Fedora docs. Anyone have thoughts/suggestions/etc? :) Regards and best wishes, Justin Clift -- Salasaga - Open Source eLearning IDE http://www.salasaga.org

2 1