May 2010 - Users - libvirt List Archives

[libvirt-users] Better error codes when stopping a VM that is already stopped
by Guido Winkelmann 30 Jun '10

30 Jun '10

Hi, Sometimes when stopping a virtual domain using virDomainDestroy(), I come across a domain that is already stopped. (For example when someone already stopped the domain manually using virsh or because the guest OS issued a shutdown.) This is a special case that I absolutely need to catch and handle. Unfortunately, when this happens, and I call virGetLastError() afterwards, I always just get the error code VIR_ERR_OPERATION_INVALID, which doesn't look very precise to me... Two questions about this: Is the domain not running at that moment the only possible condition that can trigger this particular error code when calling virDomainDestroy()? Can't we have a more obvious/precise error code in that case, like VIR_ERR_DOMAIN_NOT_RUNNING or something like that? Guido Winkelmann

4 6

[libvirt-users] Working example of logical storage pool and volume creation?
by Justin Clift 02 Jun '10

02 Jun '10

Hi all, Does anyone have a working example of creation of a logical storage pool and volume? I'm hitting a wall getting logical volumes to work on RHEL 6 beta. There's a single drive I'm trying to setup (sdc) as a libvirt managed logical storage pool, but all volume creation on it fails. Here's what I'm finding so far: Prior to any storage pool work, only the host boot drive (sda) is in lvm. # vgs VG #PV #LV #SN Attr VSize VFree vg_host2 1 3 0 wz--n- 465.27g 0 # pvs PV VG Fmt Attr PSize PFree /dev/sda2 vg_host2 lvm2 a- 465.27g 0 # ls -la /dev/mapper total 0 drwxr-xr-x. 2 root root 120 Jun 1 03:12 . drwxrwxrwt. 21 root root 3920 Jun 1 03:12 .. crw-rw----. 1 root root 10, 58 May 31 14:08 control lrwxrwxrwx. 1 root root 7 May 31 14:10 vg_host2-lv_home -> ../dm-7 lrwxrwxrwx. 1 root root 7 May 31 14:08 vg_host2-lv_root -> ../dm-0 lrwxrwxrwx. 1 root root 7 May 31 14:10 vg_host2-lv_swap -> ../dm-8 # Creation of a logical storage pool, using sdc: # virsh pool-define-as guest_images_lvm logical - - /dev/sdc libvirt_lvm /dev/mapper Pool guest_images_lvm defined # virsh pool-info guest_images_lvm Name: guest_images_lvm UUID: 17398035-8000-7252-ab25-58db95aed130 State: inactive # virsh pool-build guest_images_lvm Pool guest_images_lvm built # The physical device and volume group are seen in lvm: # pvs PV VG Fmt Attr PSize PFree /dev/sda2 vg_host2 lvm2 a- 465.27g 0 /dev/sdc libvirt_lvm lvm2 a- 465.76g 465.76g # vgs VG #PV #LV #SN Attr VSize VFree libvirt_lvm 1 0 0 wz--n- 465.76g 465.76g vg_host2 1 3 0 wz--n- 465.27g 0 # Starting the logical pool works: # virsh pool-start guest_images_lvm Pool guest_images_lvm started # Creation of volumes fail: # virsh vol-create-as guest_images_lvm volume1 8G error: Failed to create vol volume1 error: internal error '/sbin/lvcreate --name volume1 -L 8388608K /dev/mapper' exited with non-zero status 3 and signal 0: Volume group name expected (no slash) Run `lvcreate --help' for more information. # So, it looked like "/dev/mapper" in the pool-define-as command should have instead been the LVM volume group again. That doesn't work either though: # virsh pool-define-as guest_images_lvm logical - - /dev/sdc libvirt_lvm libvirt_lvm Pool guest_images_lvm defined # virsh pool-build guest_images_lvm Pool guest_images_lvm built # virsh pool-start guest_images_lvm Pool guest_images_lvm started # virsh vol-create-as guest_images_lvm volume1 8G error: Failed to create vol volume1 error: cannot read path 'libvirt_lvm/volume1': No such file or directory Looking in the libvirtd logs for this last attempt, the volume was actually successfully created. But libvirtd then checks for it's presence (wrongly!), doesn't find it, and then lvremove's it: 03:01:24.000: debug : virRunWithHook:832 : /sbin/lvcreate --name volume1 -L 8388608K libvirt_lvm ... 03:01:24.223: debug : virRunWithHook:848 : Command stdout: Logical volume "volume1" created ... 03:01:24.223: error : virStorageBackendLogicalCreateVol:616 : cannot read path 'libvirt_lvm/volume1': No such file or directory ... 03:01:24.223: debug : virRunWithHook:832 : /sbin/lvremove -f libvirt_lvm/volume1 Guessing I'm either doing something wrong, or hitting bugs, or ? Anyone have ideas/suggestions/etc? Regards and best wishes, Justin Clift -- Salasaga - Open Source eLearning IDE http://www.salasaga.org

1 1

[libvirt-users] Mount options for fs type storage pools?
by Justin Clift 02 Jun '10

02 Jun '10

Hi all, Presently writing up "Storage Pool" docs for the Fedora Virtualisation Guide. (using both virsh and virt-manager) Having trouble finding out how to specify mount options for fs type storage pools. For example: # virsh pool-define-as guest_images_fs fs - - /dev/sdc1 - "/guest_images" Pool guest_images_fs defined # virsh pool-build guest_images_fs Pool guest_images_fs built # virsh pool-start guest_images_fs Pool guest_images_fs started # mount | grep guest_images /dev/sdc1 on /guest_images type ext4 (rw) # Notice that /guest_images is mounted with default options. Nothing wrong with that, but lets say the SysAdmin has a need to pass specific mount options. (ie journal_checksum, or noatime, or nodiratime, etc) Checked the virStorageBackendFileSystemMount() function in present git master, and it doesn't look like mount options are considered for anything other than Gluster? How to do it, or is it best to not use libvirt to manage the mounting of this type of storage when options are needed? Regards and best wishes, Justin Clift -- Salasaga - Open Source eLearning IDE http://www.salasaga.org

2 2

[libvirt-users] Host should not touch passed-through USB device, how?
by Andreas Ntaflos 01 Jun '10

01 Jun '10

Hello list, this may be off topic but it is closely related to my Libvirt adventures so maybe you can help me. I want to pass a USB device (a USB Huawei E620 modem) through to a KVM guest. Pass-through in general works fine and the modem is visible in the guest. However, the host still sees it and sometimes tries to access it (such modems seem to also be USB storage devices). This naturally leads to conflicts and I believe this crashed the guest very badly yesterday. So I am looking for a way to make sure the host passes through the USB device (in its entirety) without trying to touch it and claim it as its own. Preferably without disabling the USB storage subsystem completely on the host. Any ideas? Is there a best practice for such things in the Libvirt world? Or any other world? Thanks! Andreas -- Andreas Ntaflos GPG Fingerprint: 6234 2E8E 5C81 C6CB E5EC 7E65 397C E2A8 090C A9B4

3 3

[libvirt-users] Qemu driver not supporting VIR_DOMAIN_DEVICE_MODIFY_CONFIG yet?
by Guido Winkelmann 01 Jun '10

01 Jun '10

Hi, According to the docs on http://libvirt.org/html/libvirt-libvirt.html#virDomainDetachDeviceFlags I ought to be able to do virDomainDetachDeviceFlags( <domain>, <devicexml>, VIR_DOMAIN_DEVICE_MODIFY_LIVE | VIR_DOMAIN_DEVICE_MODIFY_CONFIG); to detach a volume both from a currently running VM instance and permanently from the defined config file. If I actually try that, though, I'm getting this error: libvir: QEMU error : Requested operation is not valid: cannot modify the persistent configuration of a domain The same thing happens whether or not the domain is running and also happens if I leave out the "VIR_DOMAIN_DEVICE_MODIFY_LIVE" part. Does the qemu driver simply not support VIR_DOMAIN_DEVICE_MODIFY_CONFIG yet (libvirt 0.8.1), or is there some other problem here? Guido

2 1

[libvirt-users] Monitoring KVM Machine Uptimes
by Karsten Knese 01 Jun '10

01 Jun '10

Hey folks. Currently, I am working on an accounting-solution for Private Cloud with Eucalyptus as my bachelor-thesis. Eucalyptus bases on libvirt providing the KVM machines. Part of the thesis is kind of measurement for the uptimes of each instance in the cloud. As Eucalyptus is based on libvirt, i want to know if there is kind of tool already setup for monitoring a vm in terms of uptimes, cputimes and disk i/o. If not, can anyone give me a hint, how i could grep these facts? I found a tool called "virt-top" which monitors the vm, but for me there is to much overhead. Thanks for you support. cheers, Karsten

2 1

[libvirt-users] ballooning option xml
by udo1 01 Jun '10

01 Jun '10

Hi everybody, I need the ballooning option with kvm using libvirt xml file and virsh (virsh create domain.xml) kvm line must be appended with the -balloon virtio switch to enable ballooning. (can be disabled if there no option or with -balloon none ) I currently using Ubuntu Lucid How can I define this in the domain xml file, or what need I to do to get this option? is there any option like enabling with <balloon>virtio</balloon> or disabling with <balloon>none</balloon> Thanks in Advance Udo Attila Fischer

3 6

[libvirt-users] Update Libvirt
by Rafael Braga 31 May '10

31 May '10

Hi, I installed the opensuse 11.0 and libvirt version installed is 0.4. But, Nimbus need the version 0.6.3. How do I upgrade? I can't by YAST... thanks! -- Rafael Braga

2 2

[libvirt-users] monitoring kvm machines with libvirt
by Karsten Knese 31 May '10

31 May '10

Hey folks. Currently, i am working on my bachelorthesis. Subject of this thesis is to setup an accounting-solution for the Private Cloud with Eucalyptus. Eucalyptus works with libvirt providing kvm instances. my question is, if i can grep facts like general uptimes, cpu-peaks and disk i/o of each vm with libvirt-tools?! I just found one called virt-top, which monitors each vm, but unfortunately it produces to much overhead. If you should be no tool, can anybody give me a hint, where i best look at to grep these facts? First try is to grep the uptime of each vm by writing a little deamon, checking one time per minute via virsh, if the vm is in state running. But i don't know, if this is best. thanks for your support! cheers, Karsten

1 0

[libvirt-users] VM parameters best practice?
by Guido Winkelmann 31 May '10

31 May '10

Hi, When creating many virtual machines from a template (disk image), you often need to set some different parameters for each vm. Those usually include root passwords and/or ssh public keys for login, ssh Host keys (so that a) not all the VMs have the same host key and b) the user can know the real host key for his vm before connecting for the first time) plus any number of user-defined parameters. IP addresses and hostnames can be set via DHCP. Obviously, some of those parameters need to be treated as confidential, so you cannot just grab them from some FTP server somewhere. In fact, as long as you haven't set any authentication credentials (like an SSL Client Certificate for example) that the vm can use to uniquely and securely identify itself to some server, you cannot securely get any of these parameters over the network. How do people usually solve that sort of problem? What is considered best practice here? Currently, my own plans are to write these parameters to a floppy image, use Unix file permissions to make sure nobody else can read the image, and then hook up that image to a virtual floppy drive in the respective vm. A shell script in the vm template can then read and apply all those parameter at some early stage during the very first boot. Unfortunately, this means mounting the image in the host system, writing the parameters and then unmounting it again - which seems sort of error prone to me in a fully automated environment. (What if two vms are being created simultaneously? What if the script crashes or hangs somewhere between mounting and unmounting the image?) Regards Guido Winkelmann

1 0

[libvirt-users] Cannot create qcow2 images with libvirt 0.8.1
by Guido Winkelmann 31 May '10

31 May '10

Hi, After upgrading to libvirt 0.8.1, I can no longer create empty volumes with no backing store and an explicit format of qcow2. This XML volume definition: <volume> <name>testserverb-data2.img</name> <allocation>0</allocation> <capacity units='G'>20</capacity> <target> <format type='qcow2'/> <permissions> <owner>107</owner> <group>107</group> <mode>0744</mode> </permissions> </target> </volume> yields the error "Cannot run /usr/bin/qemu-img to create /data/testserverb- data.img: Invalid argument" In /var/log/messages on the qemu host I can find the following snippet after that: error : virRunWithHook:933 : internal error '/usr/bin/qemu-img create -f qcow2 /data/testserverb-data.img 0K' exited with non-zero status 1 and signal 0: qemu-img: Image creation needs a size parameter#012 error : virStorageBackendCreateExecCommand:537 : Cannot run /usr/bin/qemu-img to create /data/testserverb-data.img: Invalid argument error : qemudClientReadBuf:1653 : gnutls_record_recv: A TLS packet with unexpected length was received. If I leave out the <format> element and/or specify a <backingStore> element for the volume, it works again. (I suppose the backing store makes it work because in that case qemu-img does not need a size parameter...) The host system is a vanilla Fedora Core 12 system (aside from the upgraded libvirt). Regards, Guido Winkelmann

2 3

[libvirt-users] Hot-add interface crashes libvirt
by steven765＠yahoo.com 27 May '10

27 May '10

Hi Everyone, After upgrading FC12 and now FC13 whenever I try to attach a network device using attach-device it promptly crashes libvirtd. Is there an option that changed in the network interface xml spec I missed? I really need this feature or at least the ability to connect and disconnect interfaces if that's possible? I've tried defining my network card in several ways: First the one that worked previously: <interface type="network"> <mac address="52:54:00:5c:fe:a8"/> <source network="fedora"/> </interface> Second copying what appeared in the xml after using the add hardware through the VMM: <interface type='network'> <mac address='52:54:00:5c:fe:a8'/> <source network='fedora'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x0'/> </interface> Both ways fail now. Please help. Thanks, Steve

2 1

[libvirt-users] How to obtain rights to extend libvirt documentation pages
by Марк Коренберг 25 May '10

25 May '10

For example, http://libvirt.org/formatdomain.html#elementsDisks does not contain <serial>xxxx</serial> tag. (https://bugzilla.redhat.com/show_bug.cgi?id=593266) I wrote to community (users and developers). No one has added this information to the site... -- Segmentation fault

3 2

[libvirt-users] NAT-based network 2x-4x faster than routed network?
by Kenni Lund 22 May '10

22 May '10

Hi Is it expected behaviour that a libvirt NAT-based network is faster than a libvirt routed network? I would guess that the routed network would be the fastest one in all cases, but I'm seeing the opposite in my setup. On a 100mbit internet connection, the routed network tops at 4mbyte/sec, while the nat-based network goes all the way to 10mbyte/sec. I've spend the last 6 hours on trying to figure out the cause of this, changing settings, reinstalling the system, but the problem persists and I can't find the cause of it. It is a clean and fairly simple setup: - Clean installation of CentOS 5.5 - Installation of kvm+libvirt+virt-manager+xorg - iptables service disabled - 1 public IP for the host (xxx.yyy.zzz.215) on a /32 subnet with xxx.yyy.zzz.193 as gateway - 3 public IP for the routed network (xxx.yyy.zzz.251-253) on a /26 subnet with the same gateway as the host - 1 virtual machine with either the NAT or the routed libvirt network assigned I have no networking connection errors or similar, it works perfectly with both the NAT and the routed network, but the routed network is just 2-4 times slower than the NAT one. Do you have any suggestions of what the problem might be? When I look at the iptables rules generated by libvirt, and the routing table, everything looks fine AFAICT. Traceroute from guest doesn't reveal anything either. Ping times from the two networks are similar, so the main issue seems to be throughput. Any help or pointers to what I should look at, is highly appreciated... :) I've copy pasted some of my relevant system configuration into pastebin: http://pastebin.com/jtTrHLqA Thank you, Kenni

1 0

[libvirt-users] Update of libvirt website required. not all information about XML tags
by Марк Коренберг 21 May '10

21 May '10

The problem desribed here: https://bugzilla.redhat.com/show_bug.cgi?id=593266 Please update http://www.libvirt.org/formatdomain.html and describe NEW xml tags, like <serial>xxxx</serial> for hard drive and all cpu tags, other if any. -- Segmentation fault

4 8

[libvirt-users] ctrl+alt+2 disabled?
by Ralf Hornik Mailings 19 May '10

19 May '10

Dear list, I need to attach pci devices using qemu command interface but when I use the vnc connection started using virsh, the key combination (or qemu interface) seems to be disabled. How can I (re)enable it? Thanks and best regards Ralf

2 3

[libvirt-users] Can't get audio device through virt-manager
by Arthur Pemberton 18 May '10

18 May '10

virt-manager is not asking qemu to produce any sound device based on the kvm-qemu I see with 'ps aux' https://bugzilla.redhat.com/show_bug.cgi?id=592671 -- Fedora 11 (www.pembo13.com)

1 0

[libvirt-users] learn XML <network/>
by Michael Löblich 18 May '10

18 May '10

Can please anybody write a XML-description for the diagram shown on http://libvirt.org/archnetwork.html Michael

3 3

[libvirt-users] virsh domxml-from-native not working
by Ralf Hornik Mailings 17 May '10

17 May '10

Dear list, when I try to convert the (working) command /usr/bin/qemu-system-x86_64 \ -drive file=/dev/xen_vol/xen_ca,cache=none,if=virtio,boot=on \ -net nic,model=virtio,macaddr=00:16:3E:04:F5:6D \ -net tap,ifname=tap1 -m 256 -name ca \ -vnc 10.0.0.100:4 -daemonize -k de into an xml file: <domain type='qemu'> <name>ca</name> <uuid>dfef3884-d7a9-f6cf-5204-1615963cc0e8</uuid> <memory>262144</memory> <currentMemory>262144</currentMemory> <vcpu>1</vcpu> <os> <type arch='x86_64'>hvm</type> </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='block' device='disk'> <driver cache='none'/> <source dev='/dev/xen_vol/xen_ca'/> <target dev='vda' bus='virtio'/> </disk> <interface type='ethernet'> <mac address='00:16:3e:04:f5:6d'/> <target dev='tap1'/> <model type='virtio'/> </interface> <input type='mouse' bus='ps2'/> <graphics type='vnc' port='5904' autoport='no' listen='10.0.0.100'/> <video> <model type='cirrus' vram='9216' heads='1'/> </video> </devices> </domain> Then it stops to boot while haning on "Loading Grub Stage 2". Additionally, the keyboard encoding "de" is missing. The corresponding process is: /usr/bin/qemu-system-x86_64 -S -M pc-0.12 -no-kvm -m 256 \ -smp 1,sockets=1,cores=1,threads=1 \ -name ca -uuid dfef3884-d7a9-f6cf-5204-1615963cc0e8 \ -nodefaults \ -chardev socket,id=monitor,path=/usr/local/libvirt/var/lib/libvirt/qemu/ca.monitor,server,nowait \ -mon chardev=monitor,mode=readline -rtc base=utc -boot c \ -drive file=/dev/xen_vol/xen_ca,if=none,id=drive-virtio-disk0,boot=on,cache=none \ -device virtio-blk-pci,bus=pci.0,addr=0x4,drive=drive-virtio-disk0,id=virtio-disk0 \ -device virtio-net-pci,vlan=0,id=net0,mac=00:16:3e:04:f5:6d,bus=pci.0,addr=0x5 \ -net tap,ifname=tap1,vlan=0,name=hostnet0 -usb -vnc 10.0.0.100:4 \ -vga cirrus -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 Any Ideas, hats going wrong? Thank yout and best regards Ralf Hornik

1 0

[libvirt-users] Sharing same storage on two diffrent KVM hosts
by shy Perets 17 May '10

17 May '10

Hello , I whould like your suggestion regarding the following problem : I got 2 host which are running under RHEL5.4 hypervizor . both of the hosts disks is a logical volume .( Each for each ) under the same VG . I want to add storage ( another logical volume ) which will be availble for both of the host as local disk . ( Share storage ) Thanks .

1 0

[libvirt-users] Can not restore domain from a shared state file
by Zhang Qian 16 May '10

16 May '10

Hi, I have two KVM host: h1 and h2, both of them mount an NFS directory as a shared storage. I can save (virsh save <domain> <file>) a domain in h1 to a state file in the shared storage successfully, but failed to restore it from h2 with the following error message: # virsh restore testRes.dat error: Failed to restore domain from testRes.dat error: operation failed: failed to start VM I can always restore it from h1, but sometimes works for h2 (wait for a while, then "virsh restore" command may succeed in h2). I guess the state file generated by "virsh save" command is not intact from h2 point view, may be cause by the cache of NFS server? Any suggestions will be really appreciated, thanks in advance! Regards, Qian

2 3

[libvirt-users] XML-Created domain vanishes after shutdown
by Alexander Topolanek 16 May '10

16 May '10

Hi, im migrating existing kvm-images into libvirt domains, by creating a domain in virsh from a xml file. The domain starts, is working fine, but after shutdown it is lost, and I have to recreate it. BTW, can I reset the domain count somewhere? Any ideas? thx and kind regards Alexander virsh # version Compiled against library: libvir 0.4.6 Using library: libvir 0.4.6 Using API: QEMU 0.4.6 Running hypervisor: QEMU 0.9.1 This is my xml file: <domain type='qemu' id='2'> <name>munin</name> <uuid>6bb2995a-5443-2ddb-67fe-d69a9814ece4</uuid> <memory>262144</memory> <currentMemory>524288</currentMemory> <vcpu>2</vcpu> <os> <type arch='x86_64' machine='pc'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> <pae/> </features> <clock offset='utc'/> <on_poweroff>preserve</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/bin/kvm</emulator> <disk type='file' device='disk'> <source file='/var/kvm/virsh/munin.img'/> <target dev='hda' bus='ide'/> </disk> <interface type='bridge'> <mac address='00:30:1c:45:40:02'/> <source bridge='br0'/> <target dev='vnet0'/> <model type='virtio'/> </interface> <input type='mouse' bus='ps2'/> <graphics type='vnc' port='5900' autoport='yes' listen='127.0.0.1' keymap='de'/> </devices> </domain>

2 1

[libvirt-users] Red Hat Bugzilla – Bug 592671 : can't use any sound device from virt-manager
by Arthur Pemberton 16 May '10

16 May '10

url : https://bugzilla.redhat.com/show_bug.cgi?id=592671 basically I'm not getting any sound devices when invoking from virt-manager at least. -- Fedora 11 (www.pembo13.com)

1 0

[libvirt-users] How to eject cdrom by libvirt API
by Zhang Qian 13 May '10

13 May '10

Hi, I want to eject cdrom by calling libvirt C API, I found there is a similar feature in virt-manager: disconnect cdrom from the related iso file, I am wondering how to achieve this by libvirt API. Thanks! Regards, Qian

1 0

[libvirt-users] remote libvirt vbox
by angelo alfano 11 May '10

11 May '10

I tried: import libvirt conn = libvirt.open("vbox+ssh://user@192.168.0.10/session") Prompted for a password for ssh connection, but returns this error: libvir: QEMU error : server closed connection Traceback (most recent call last): File "/home/angelo/workspace/test.py", line 11, in <module> conn = libvirt.open("vbox+ssh://user@192.168.0.10/session") File "/usr/lib/python2.6/dist-packages/libvirt.py", line 169, in open if ret is None:raise libvirtError('virConnectOpen() failed') libvirt.libvirtError: server closed connection

1 0