- Users - libvirt List Archives

[libvirt-users] Bug in VNC QEMU?
by Nikita A Menkovich 08 Aug '12

08 Aug '12

Hi all! Recently I found some problems after upgrading one of our virtual machines from ubuntu 10.10. to later releases. Root problem is inactive VNC terminal after booting grub. It happens because of using framebuffer at boot process. To fix this problem you should change at /etc/default/grub: GRUB_TERMINAL=console And run # update-grub # reboot This fixes problem. But I think there should be some fix to use fb in vnc without problems. I use: libvirt 0.9.11.3 QEMU emulator version 1.0 (qemu-kvm-1.0 Debian 1.0+dfsg-8~bpo60+1), Copyright (c) 2003-2008 Fabrice Bellard <graphics type='vnc' port='PORT' autoport='no' listen='IP'> <listen type='address' address='IP'/> </graphics> <video> <model type='cirrus' vram='9216' heads='1'/> <alias name='video0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </video> -- Nikita A Menkovich http://libc6.org/ JID: menkovich(a)gmail.com

1 0

[libvirt-users] Channel with virtio problem
by norazah aziz 08 Aug '12

08 Aug '12

I have configured the kvm and edit the xml as below, however there are an error when I want to run the guest. The error as follows: virsh create libvirt.xml error: Failed to create domain from libvirt-test.xml error: XML description for unknown target type for character device: virtio is not well formed or invalid Please help me to solve this problem. ############################################################ <domain type='kvm'> <name>i-5E36098H</name> <os> <type>hvm</type> <kernel>/opt/eucalyptus-data/suria/i-5E36098H/kernel</kernel> <initrd>/opt/eucalyptus-data/suria/i-5E36098H/ramdisk</initrd> <cmdline>root=LABEL=root console=ttyS0</cmdline> </os> <features> <acpi/> </features> <memory>13107200</memory> <vcpu>1</vcpu> <devices> <emulator>/usr/bin/kvm</emulator> <disk type='file'> <source file='/opt/eucalyptus-data/suria/i-5E36098H/disk'/> <target dev='vda' bus='virtio'/> </disk> <interface type='network'> <source network='default'/> <mac address='d0:0d:45:D2:08:01'/> <model type='virtio'/> </interface> <controller type='ide' index='0'/> <controller type='virtio-serial' index='0' ports='16' vectors='4'/> <controller type='virtio-serial' index='1'> <address type='pci' domain='0x0000' bus='0x00' slot='0x0a' function='0x0'/> </controller> <parallel type='pty'> <source path='/dev/pts/2'/> <target port='0'/> </parallel> <serial type='pty'> <source path='/dev/pts/3'/>  <target type='virtio' port='0'/> </serial> <console type='pty'> <source path='/dev/pts/4'/> <target type='virtio' port='0'/> </console> <channel type='pty'> <source mode='bind' path='/var/lib/libvirt/qemu/va-i5E36098H-virtio.sock'/> <target type='virtio' name='org.qemu.guest_agent.0'/> <address type='virtio-serial' controller='0' bus='0' port='0'/> </channel> <graphics type='vnc' port='-1' autoport='yes' keymap='en-us'/> </devices> ########################################################################

1 0

[libvirt-users] Authentication via SASL and LDAP?
by Andrew Martin 08 Aug '12

08 Aug '12

Hello, I've recently configured a new virtual machine host running Ubuntu 12.04 server with libvirt and KVM. I am configuring WebVirtMgr ( https://github.com/retspen/webvirtmgr/ ) for users to manage machines via a web interface. This requires access to the host using qemu+tcp, which I have configured as follows: /etc/default/libvirt-bin: start_libvirtd="yes" libvirtd_opts="-d -l" /etc/libvirt/libvirtd.conf: listen_tls = 0 listen_tcp = 1 unix_sock_group = "libvirtd" unix_sock_rw_perms = "0770" auth_unix_ro = "none" auth_unix_rw = "none" auth_tcp = "sasl" The libvirt documentation ( http://libvirt.org/auth.html ) does not specify if it is possible to use SASL with a different authentication method other than DIGEST-MD5. I would like to authenticate users via LDAP - is this possible? I configured LDAP authentication via SASL as follows: /etc/sasl2/libvirt.conf pwcheck_method: saslauthd mech_list: PLAIN LOGIN log_level: 5 saslauthd_path: /var/run/saslauthd/mux auxprop_plugin: ldap /etc/saslauthd.conf ldap_servers: ldap://ldap_ip_addr:389/ ldap_search_base: ou=People,dc=x-es,dc=com ldap_auth_method: none ldap_filter: uid=%u ldap_version: 3 /etc/default/saslauthd START=yes DESC="SASL Authentication Daemon" NAME="saslauthd" MECHANISMS="ldap" MECH_OPTIONS="" THREADS=5 OPTIONS="-O /etc/saslauthd.conf -c -m /var/run/saslauthd -r" Testing the configuration works: # testsaslauthd -u myuser -p mypass 0: OK "Success." However, if I attempt to connect over the libvirt TCP connection I am denied: virsh -c qemu+tcp://my_vm_host/system nodeinfo error: authentication failed: Failed to start SASL negotiation: -4 (SASL(-4): no mechanism available: ) error: failed to connect to the hypervisor Do you know what is incorrect in my libvirt config, or is it not possible to authenticate libvirt via SASL+LDAP? If not, is there somewhere that documents all of the supported mechanisms? Thanks, Andrew Martin

1 0

[libvirt-users] using RBD with libvirt 0.9.13
by Jeff Strunk 07 Aug '12

07 Aug '12

I'm having some trouble creating KVM domains with RBD block devices using virsh. I've managed to get virsh to define the domain, but it gives an error when trying to start the domain: error: Failed to start domain test0 error: internal error process exited while connecting to monitor: char device redirected to /dev/pts/3 kvm: -drive file=rbd:rbd/test0,if=none,id=drive-virtio-disk0,format=raw: error connecting kvm: -drive file=rbd:rbd/test0,if=none,id=drive-virtio-disk0,format=raw: could not open disk image rbd:rbd/test0: No such file or directory >From libvirt's QEMU log: 2012-08-06 18:36:57.788+0000: starting up LC_ALL=C PATH=/usr/local/sbin:/usr/local/bin:/usr/bin:/usr/sbin:/sbin:/bin QEMU_AUDIO_DRV=none /usr/bin/kvm -name test0 -S -M pc-1.0 -enable-kvm -m 1024 -smp 1,sockets=1,cores=1,threads=1 -uuid 3e727624-f336-d674-4e0c-234641fc9987 -nodefconfig -nodefaults -chardev socket,id=charmonitor,path=/var/lib/libvirt/qemu/test0.monitor,server,nowait -mon chardev=charmonitor,id=monitor,mode=control -rtc base=utc -no-shutdown -device piix3-usb-uhci,id=usb,bus=pci.0,addr=0x1.0x2 -drive file=rbd:rbd/test0,if=none,id=drive-virtio-disk0,format=raw -device virtio-blk-pci,scsi=off,bus=pci.0,addr=0x4,drive=drive-virtio-disk0,id=virtio-disk0,bootindex=1 -drive if=none,id=drive-ide0-1-0,readonly=on,format=raw -device ide-cd,bus=ide.1,unit=0,drive=drive-ide0-1-0,id=ide0-1-0 -netdev tap,fd=19,id=hostnet0 -device virtio-net-pci,netdev=hostnet0,id=net0,mac=52:54:00:8c:e4:92,bus=pci.0,addr=0x3 -chardev pty,id=charserial0 -device isa-serial,chardev=charserial0,id=serial0 -vnc 127.0.0.1:0 -vga cirrus -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x5 Domain id=5 is tainted: high-privileges char device redirected to /dev/pts/2 kvm: -drive file=rbd:rbd/test0,if=none,id=drive-virtio-disk0,format=raw: error connecting kvm: -drive file=rbd:rbd/test0,if=none,id=drive-virtio-disk0,format=raw: could not open disk image rbd:rbd/test0: No such file or directory 2012-08-06 18:36:58.353+0000: shutting down I can, however, successfully run a KVM virtual machine with an RBD block device with the following command: kvm -drive file=rbd:rbd/test1,id=drive-virtio-disk0,format=raw -cdrom /var/lib/libvirt/images/ubuntu-12.04-server-amd64.iso kvm will boot with the "-drive file=rbd:rbd/test0,if=none,id=drive-virtio-disk0,format=raw" options, but the disk is not detected unless "if=none" is removed. Has anyone else figured this out? Thank you, Jeff Here is my rbd pool definition: ######### BEGIN <pool type="rbd"> <name>rbd</name> <source> <name>rbd</name> <host name='thinkmate3' port='6789'/> <host name='thinkmate4' port='6789'/> <auth username='admin' type='ceph'> <secret uuid='f0836fcb-9c05-5ccf-653a-eb1d0308e247'/> </auth> </source> </pool> ######### END My rbd volumes: # virsh vol-list rbd Name Path ----------------------------------------- test0 rbd/test0 test1 rbd/test1 test2 rbd/test2 Here is my domain definition: ######### BEGIN <domain type='kvm'> <name>test0</name> <uuid>3e727624-f336-d674-4e0c-234641fc9987</uuid> <memory unit='KiB'>1048576</memory> <currentMemory unit='KiB'>1048576</currentMemory> <vcpu placement='static'>1</vcpu> <os> <type arch='x86_64' machine='pc-1.0'>hvm</type> <boot dev='hd'/> </os> <features> <acpi/> <apic/> <pae/> </features> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/bin/kvm</emulator> <disk type='network' device='disk'> <driver name='qemu' type='raw'/> <source protocol='rbd' name='rbd/test0'/> <target dev='vda' bus='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </disk> <disk type='block' device='cdrom'> <driver name='qemu' type='raw'/> <target dev='hdc' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='1' target='0' unit='0'/> </disk> <controller type='ide' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <controller type='usb' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x2'/> </controller> <interface type='network'> <mac address='52:54:00:8c:e4:92'/> <source network='default'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <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'/> <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> ######## END

2 3

[libvirt-users] Opteron_G4 CPU under libvirt 0.9.12
by Michael Beham 06 Aug '12

06 Aug '12

Hi, I´m using libvirt version 0.9.12 under Debian Squeeze with and AMD Opteron 4280. Executing the virsh capabilities command only show me the following flags: <cpu> <arch>x86_64</arch> <model>Opteron_G4</model> <vendor>AMD</vendor> <topology sockets='1' cores='8' threads='2'/> <feature name='nodeid_msr'/> <feature name='wdt'/> <feature name='skinit'/> <feature name='ibs'/> <feature name='osvw'/> <feature name='cr8legacy'/> <feature name='extapic'/> <feature name='cmp_legacy'/> <feature name='fxsr_opt'/> <feature name='mmxext'/> <feature name='osxsave'/> <feature name='monitor'/> <feature name='ht'/> <feature name='vme'/> </cpu> Should not there be more flags supported by the CPU, for example the SVM-Flag for virtualization because executing cat /proc/cpuinfo gives me quite a lot of more flags and also the SVM Flag. Cheers, Mike

2 1

[libvirt-users] CPU Flags libvirt 0.9.12
by Michael Beham 04 Aug '12

04 Aug '12

Hi, I?m using libvirt version 0.9.12 under Debian Squeeze with and AMD Opteron 4280. Executing the virsh capabilities command only show me the following flags: <cpu> <arch>x86_64</arch> <model>Opteron_G4</model> <vendor>AMD</vendor> <topology sockets='1' cores='8' threads='2'/> <feature name='nodeid_msr'/> <feature name='wdt'/> <feature name='skinit'/> <feature name='ibs'/> <feature name='osvw'/> <feature name='cr8legacy'/> <feature name='extapic'/> <feature name='cmp_legacy'/> <feature name='fxsr_opt'/> <feature name='mmxext'/> <feature name='osxsave'/> <feature name='monitor'/> <feature name='ht'/> <feature name='vme'/> </cpu> Should not there be more flags supported by the CPU, for example the SVM-Flag for virtualization because executing cat /proc/cpuinfo gives me quite a lot of more flags and also the SVM Flag. Cheers, Mike

2 2

[libvirt-users] Determine state of the guest OS/kernel from host
by Harish Patil 04 Aug '12

04 Aug '12

Hello, Is there a way to determine state of the qemu/kvm guest OS (rebooting/working/hung etc) from the host. I'm not referring to the qemu domain state itself Thanks in advance. >harish

3 5

[libvirt-users] Unable to load library 'virt'
by zhijun liu 04 Aug '12

04 Aug '12

Hi,all now ,an old problem I really cann't resolved ,in my window7(64bit) use libvirt-0.4.7.jar to conn server System.setProperty("jna.library.path", "D:/Program Files/Libvirt/bin"); conn = new Connect("qemu+tcp://192.168.1.106/system",true); the exception is Exception in thread "main" java.lang.UnsatisfiedLinkError: Unable to load library 'virt': ÕҲ»µ½ָ¶¨ PS:in cmd line i want to connect sever is no problem ,like this: virsh -c qemu+tcp://192.168.1.106/system, it's ok thanks in advance!!!

1 0

Re: [libvirt-users] Excluding RAW disk from snapshot
by Trey Dockendorf 04 Aug '12

04 Aug '12

On Thu, Aug 2, 2012 at 11:58 AM, Eric Blake <eblake(a)redhat.com> wrote: > On 08/02/2012 10:53 AM, Trey Dockendorf wrote: > > Did you mean for this to go to the list? Others might benefit from the > answer. > >>>> >>>> <domainsnapshot> >>>> <name>20120802</name> >>>> <description>oVirt 3.1 upgrade</description> >>>> <disks> >>>> <disk name='vda' snapshot='yes' /> >>>> <disk name='vdb' snapshot='no' /> >>>> </disks> >>>> </domainsnapshot> >>>> >>>> When I try the snapshot I get this error... >>>> > >> Here's the new XML >> >> <domainsnapshot> >> <name>20120802</name> >> <description>oVirt 3.1 upgrade</description> >> <disks> >> <disk name='vdb' snapshot='no' /> >> </disks> >> </domainsnapshot> >> > > Ah, right - snapshot is a tri-state: 'no', 'internal', and 'external', > so 'yes' is not a valid state (although currently only 2 of the 3 states > are supported for qemu: 'no' and 'external'). > >> Now I get this error >> >> >> # virsh snapshot-create --disk-only dc-vmarchitect dc-vmarchitect_ss.xml >> error: operation failed: Failed to take snapshot: unknown command: >> 'snapshot_blkdev' > > That's a sign of your qemu being too old to support disk snapshots > (older qemu only supports system checkpoint internal snapshots). > >> >> I do not know if this could be an issue, but after this wasn't working >> with libvirt-0.9.4, I did "yum update qemu-img qemu-kvm libvirt" and >> am now at the versions listed in previous email. I did not do a full >> system upgrade or reboot as this is a heavily loaded production >> system. Is this the case where the VM is running as the previous >> version of qemu-kvm ? >> >> Thanks >> - Trey >> > > -- > Eric Blake eblake(a)redhat.com +1-919-301-3266 > Libvirt virtualization library http://libvirt.org > Replying back to list, accidently missed reply-all. Could it be the case that since I upgraded libvirt and qemu while all the VMs were running, that the VM has to be restarted to take on the updated version's capability? If this isn't something yet supported in EL6.3, which I hope it is, is there any other way to perform a snapshot and exclude a disk aside from temporarily detaching it from the VM? Thanks - Trey

2 1

[libvirt-users] How to connect to local hpyervisor by libvirt?
by 何鑫 03 Aug '12

03 Aug '12

Hi libvirt team, I suffered some troubles. Could you please give me some advices? Our team plan to control more then 100 android emulators, however, we failed to connect to the local android emulators by libvirt. Now we have tested 2 ways. 1. Converting from QEMU args to domain XML. It's a right way to control the local QEMU built manually. But the android emulator based on QEMU modified its command a lot, so we cancelled the way. 2. Using libvirt hypervisor driver for QEMU. Actually we think the URI ---"qemu+tcp://127.0.0.1:5555/system"(the IP and port is given by android emulator) -- is theoretically available, also, we could connect to the IP successfully by socket's program, but the result is "error: server closed connection: error: failed to connect to the hypervisor". We don't figure the reason out now. Concerning about such problmes above, I have some questions? 1.Can the libvirt connect to local hypervisors by "qemu+tcp" as described above? 2.If we can't do by "qemu+tcp", could you please give us some suggestoins? Thanks a lot!

2 1