- Users - libvirt List Archives

[libvirt-users] libvirt_lxc causes cpu 100% usage
by jj zhong 01 Nov '13

01 Nov '13

libvirt_lxc causes cpu 100% usage? libvirt version : 1.1.1 I don't know what happend, any help would be appreciated. thx top - 07:39:44 up 1 day, 2:13, 2 users, load average: 4.72, 4.59, 4.51 Tasks: 243 total, 4 running, 239 sleeping, 0 stopped, 0 zombie %Cpu0 : 33.1 us, 1.3 sy, 0.0 ni, 64.6 id, 0.7 wa, 0.0 hi, 0.3 si, 0.0 st %Cpu1 : 41.1 us, 3.0 sy, 0.0 ni, 55.3 id, 0.7 wa, 0.0 hi, 0.0 si, 0.0 st %Cpu2 : 18.0 us, 2.0 sy, 0.0 ni, 79.7 id, 0.0 wa, 0.0 hi, 0.3 si, 0.0 st %Cpu3 : 94.8 us, 5.2 sy, 0.0 ni, 0.0 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st %Cpu4 : 55.7 us, 7.5 sy, 0.0 ni, 36.7 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st %Cpu5 : 66.1 us, 4.3 sy, 0.0 ni, 29.6 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st %Cpu6 : 16.8 us, 1.0 sy, 0.0 ni, 82.2 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st %Cpu7 : 7.2 us, 0.7 sy, 0.0 ni, 92.1 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st KiB Mem: 1972112 total, 1894912 used, 77200 free, 236576 buffers KiB Swap: 10485744 total, 0 used, 10485744 free, 574544 cached PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 15750 root 20 0 10448 2080 1216 R 100.3 0.1 1322:38 libvirt_lxc 27517 root 20 0 10448 2032 1200 R 99.3 0.1 1273:44 libvirt_lxc 25250 root 20 0 10480 2048 1200 R 99.0 0.1 1277:31 libvirt_lxc ....

4 7

[libvirt-users] upstart script for virtlockd
by NEVEU Stephane 31 Oct '13

31 Oct '13

Hi all, Trying to test libvirt 1.1.3 with virtlockd locking my qcow2 images on a NFS storage between two kvm hosts. ./configure ... --with-init-script=upstart Libvirtd upstart script is actually well generated but I can't see nothing about virtlockd... or am I blind ? :) Nevertheless, running virtlockd -d && service libvirtd restart works fine. Am I wrong thinking that editing qemu-lockd.conf with my needs and qemu.conf with lock_manager="lockd" was enough ? Thank you for your help.

2 3

[libvirt-users] libvirt_lxc: SELinux MCS
by Matteo Piccinini 31 Oct '13

31 Oct '13

Hello list, my name is Matteo, i'm new on that list. I'm working on a multitenancy platform with linux containers through libvirt on a production system with Red Hat 6.4. Every container run a separate instance of OpenSSH and Apache HTTPd and I need to give root privileges to the developers and I try to configure SELinux using svirt and MCS. I try the secmodel type dynamic and static in the xml file but it didn't work, I received the following error: error : virSecurityLabelDefParseXML:3228 : XML error: security label is missing error : virNetSocketNewConnectUNIX:566 : Failed to connect socket to '/var/run/libvirt/lxc/cntr1.sock': Connection refused I configure the followings secmodel definition and used chcon on the rootfs directory (created with yum) with "system_u:object_r:svirt_lxc_file_t:s0:c30,c50" label: <seclabel type='static' model='selinux' relabel='no'> <label>system_u:system_r:svirt_lxc_net_t:s0:c30,c50</label> </seclabel> or: <seclabel type='dynamic' model='selinux' relabel='yes'> <label>system_u:system_r:svirt_lxc_net_t:s0:c30,c50</label> </seclabel> I try to compile the last version from the master branch of git the result was always the same, the error was related to the SELinux driver not enabled. The output from "virsh -c lxc:/// capabilities" doesn't show the secmodel and doi tag like the qemu/kvm have for the lxc driver the selinux driver. How can I enable the SELinux driver for libvirt lxc in Red Hat 6.4? SELinux is in enforcing mode. Thanks in advance, Matteo

2 3

[libvirt-users] libvirt-lxc driver on armv7l
by Jan Olszak 31 Oct '13

31 Oct '13

Hi there, I'm interested in using libvirt-lxc on armv7l architecture and I have a hard time building libvirt. 1. Are there any prebuild rpms for ARM? 2. I have many dependency issues during build: error: *** Error Summary *** === the following packages failed to build due to missing build dependencies === libvirt: nothing provides hal-devel nothing provides libnl-devel nothing provides libselinux-devel nothing provides radvd nothing provides ebtables nothing provides cyrus-sasl-devel nothing provides /usr/bin/qemu-img nothing provides lvm2 nothing provides iscsi-initiator-utils nothing provides parted-devel nothing provides device-mapper-devel nothing provides libssh2-devel >= 1.3.0 nothing provides libwsman-devel >= 2.2.3 nothing provides nfs-utils nothing provides scrub Where do I define the flags that discard some of the libraries (for example "--without-hal")? Thanks! Jan

2 3

[libvirt-users] libvirt-lxc without libvirt
by Jan Olszak 31 Oct '13

31 Oct '13

Hi there! 1. Is it possible to build libvirt-lxc driver without building libvirt? How do I do it? 2. Do I understand it correctly, that I can use libvirt-lxc as a substitute of lxc? Thanks, Jan

2 1

[libvirt-users] Fwd: libvirt unsupport rbd storage pool? "missing backend for pool type 8"
by 卢惠勇 31 Oct '13

31 Oct '13

I use "virsh pool-define rbd.xml" to create a rbd storage pool,but get this error virsh pool-define /tmp/rbd.xml error: Failed to define pool from /tmp/rbd.xml error: internal error: missing backend for pool type 8 rbd.xml <pool type="rbd"> <name>cloudstack</name> <source> <name>cloudstack</name> <host name='10.20.197.22' port='6789'/> </source> </pool> this is my system information: centos6.4 2.6.32-358.14.2.el6.x86_64 libvirt-1.1.1-1.el6.x86_64 qemu-kvm-0.12.1.2-2.355.el6.2.cuttlefish.x86_64 (download from http://ceph.com/packages/qemu-kvm/centos/x86_64/ ) qemu-img-0.12.1.2-2.355.el6.2.cuttlefish.x86_64 -- ------------------------------ 金山网络运维部-卢惠勇分机:2477 手机:15018838589 qq:31302409

2 1

[libvirt-users] Upgraded to F18: now machines won't start
by Sean Darcy 30 Oct '13

30 Oct '13

We've gone back to using an older machine that used to run libvirt just fine. Updated to F18, libvirt-1.1.0-1.fc18.x86_64. VirtualBox works ok. Now I can't start any image. For one machine i get: .............. File "/usr/lib64/python2.7/site-packages/libvirt.py", line 698, in create if ret == -1: raise libvirtError ('virDomainCreate() failed', dom=self) libvirtError: internal error Cannot find suitable CPU model for given data that's whether I use what we used to use "Phenom" or the selection from allowing virt-manager to choose "Opteron G3". For another machine we get: libvirtError: internal error Process exited while reading console log output: char device redirected to /dev/pts/37 (label charserial0) failed to initialize KVM: Device or resource busy Doing some debugging, I found: 1. On boot the kvm_amd and kvm kernel modules won't/don't load. Nothing in dmesg. Modprobe does load them. lsmod | grep kvm kvm_amd 59982 0 kvm 429349 1 kvm_amd I still can't make anything work with the modules loaded. 2. When I start libvirtd, I get: Oct 30 14:56:29 sixcore systemd[1]: Starting Virtualization daemon... Oct 30 14:56:29 sixcore systemd[1]: Started Virtualization daemon. ............ Oct 30 14:56:31 sixcore kernel: [ 837.705913] kvm: enabling virtualization on CPU5 failed Oct 30 14:56:31 sixcore kernel: [ 837.705920] kvm: enabling virtualization on CPU4 failed Oct 30 14:56:31 sixcore kernel: [ 837.705928] kvm: enabling virtualization on CPU1 failed Oct 30 14:56:31 sixcore kernel: [ 837.705934] kvm: enabling virtualization on CPU2 failed Oct 30 14:56:31 sixcore kernel: [ 837.705946] kvm: enabling virtualization on CPU3 failed Oct 30 14:56:31 sixcore kernel: [ 837.745267] kvm: enabling virtualization on CPU0 failed I looked at the wiki on how to debug, and I seem to pass all the tests. virsh capabilities | grep kvm <domain type='kvm'> <emulator>/usr/bin/qemu-kvm</emulator> <domain type='kvm'> <emulator>/usr/bin/qemu-kvm</emulator> virsh dumpxml 2008R2 | grep kvm <domain type='kvm'> <emulator>/usr/bin/qemu-kvm</emulator> Am I right that some of the entries in the xml file have now changed? Is there a guide to updating them from the prior entries? And, if I try to create a new machine, virt-install just hangs: virt-install --prompt What is the name of your virtual machine? New-2008R2 How much RAM should be allocated (in megabytes)? 1200 What would you like to use as the disk (file path)? /opt/storage/virt/windows/images/2008R2.img ls -l /opt/storage/virt/windows/images/2008R2.img -rw-------. 1 root root 23571726336 Dec 8 2012 /opt/storage/virt/windows/images/2008R2.img I changed the permissions to 666. still hung. Any help appreciated. Sean

1 0

[libvirt-users] virsh esx return "error: internal error: Could not find Datacenter"
by Andrea Buzzi (abuzzi) 30 Oct '13

30 Oct '13

Hello, I’m running virsh v1.1.3 compiled on a ubuntu system with the "--with-esx=yes” support. root@kvm-189:/home/cisco# virsh --version 1.1.3 I also have a vcenter 5.5 environment that has following structure (couple of hosts at datacenter level, others part of cluster group): MIC-VC-ADMIN MIC-DC-ADMIN 10.58.6.110 10.58.6.113 MIC-CL-ADMIN 10.58.6.109 10.58.6.118 10.58.6.126 I can succesfully connect to ESXi host 10.58.6.110: root@kvm-189:/home/cisco# virsh -c esx://10.58.6.110?no_verify=1 list --all Enter username for 10.58.6.110 [root]: Enter root's password for 10.58.6.110: Id Name State ---------------------------------------------------- 250 ssi-vepc1 running - SSI-Template shut off root@kvm-189:/home/cisco# However if I try to connect to vcenter, I get the following: root@kvm-189:/home/cisco# virsh -c esx://10.58.6.110?no_verify=1\&vcenter=* Enter username for 10.58.6.110 [root]: Enter root's password for 10.58.6.110: Enter username for 10.58.6.7 [administrator]: Administrator Enter Administrator's password for 10.58.6.7: error: failed to connect to the hypervisor error: internal error: Could not find Datacenter root@kvm-189:/home/cisco# Any suggestion on how to isolate (and possibly fix) the issue ? Thanks, Andrea -- [Description: Description: Description: http://www.cisco.com/swa/i/logo.gif] Andrea Buzzi TME @ MIBU abuzzi(a)cisco.com<mailto:abuzzi@cisco.com> Phone: +39 039.6295.319 Mobile: +39 348.711.9960 Cisco Systems Italy S.r.l Via Torri Bianche, 8 20871 Vimercate (MB) Italy Cisco.com<http://www.cisco.com/global/IT/> This email may contain confidential and privileged material for the sole use of the intended recipient. Any review, use, distribution or disclosure by others is strictly prohibited. If you are not the intended recipient (or authorized to receive for the recipient), please contact the sender by reply email and delete all copies of this message. For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/index.html

1 0

[libvirt-users] A RFE for libvirt
by shyu 30 Oct '13

30 Oct '13

Hi libvirt guys, I am an newbie to libvirt. There is an requirement for libvirt: Every time I want to rename an guest by virsh command, I need to virsh edit origin_guest(edit the name,delete/edit the uuid) after that a new guest with the same configuration as origin_guest was defined, then virsh undefine origin_guest. Seems it's a little complex, so should be there an virsh command like "virsh rename origin_guest new_guest"? I know I can use virt-manager to rename an guest,but if there is an virsh command can do it, I think it's much useful. Thanks -- Regards shyu

2 2

[libvirt-users] Automatically assign static ipv4 via dhcp to new VMs
by Tim 30 Oct '13

30 Oct '13

Hi guys, I'm running a KVM/libvirt host in a datacenter and got a fancy IPv4 subnet from my provider. I'm able to assign IPs from that subnet via dhcp to my VMs: host01 ~ # virsh net-dumpxml internet <network connections='3'> <name>internet</name> <uuid>37b888cc-510f-46f1-9246-346da96222ed</uuid> <forward dev='enp5s0f0' mode='route'> <interface dev='enp5s0f0'/> </forward> <bridge name='virbr1' stp='on' delay='0'/> <mac address='52:54:00:13:39:d5'/> <ip address='XX.YY.253.2' netmask='255.255.255.224'> <dhcp> <range start='XX.YY.253.4' end='XX.YY.253.30'/> <bootp file='pxelinux.0' server='XX.YY.99.100'/> </dhcp> </ip> </network> What I would like to do: 1) Right know it seems like dnsmasq assigns ip-addresses randomly to VMs, is there any way to assign them in ascending order? 2) IPs are not statically assigned. Is it possible to add something like this automatically at the first VM-start to the xml definition?: <host mac='AA:BB:CC:DD:EE:FF' ip='192.168.122.2' name='foobar'/>

3 3