- Users - libvirt List Archives

[libvirt-users] libvirt: how to set PV-on-HVM drivers??
by zcs 01 Aug '12

01 Aug '12

Hi friend. I have a problem about using libvirt now. There is a feather about HVM driver ---PV_on_HVM_drivers, in xm configure, using flag xen_platform_pci can control the feather, but in libvirt I can not find the flags to use pv-on-HVM drivers. So I want to find the answer if the libvirt can control the modules using xml file? thanks a lot. best regards. zhong.chunshan http://wiki.xen.org/wiki/Xen_Linux_PV_on_HVM_drivers

1 0

[libvirt-users] live migration causes guest system to crash when accessing the network
by Sebastian J. Bronner 30 Jul '12

30 Jul '12

Hi all, we've been using libvirt/KVM ever since it was included in Ubuntu 10.04 LTS. At first, it wasn't possible to do live block migrations. We got used to that. And now block migration has been possible since Ubuntu 12.04 LTS. And we're excited. There's just one hitch: Many of migrations have the result that the system is successfully migrated, but will no longer run. Sometimes, it still runs but can't access the network. Sometimes it just freezes completely, and doesn't even show a kernel panic on the console. And sometimes the network still works but will crash when sending ICMP messages from within the guest. And every once in a while the guest system will continue to run. Most of our guests use virtio for both networking and storage. The guest OS's are mostly Ubuntu 12.04, but also one 11.04 and several 10.04's. A few older ones are still around, as well. The older ones don't support virtio though, so here we use the emulated intel e1000 network card and IDE storage. The guests are connected to a common bridge on the node, which provides access to the outside network. There is no NAT. Oh, and I read that different CPU capabilities (flags, etc.) can cause live migrations to fail. So I used the recommended procedure to determine the least common denominator of the CPUs on our nodes and configured all the guests to use only those CPU capabilities. After a restart, I verified that the kernel correctly identified the configured CPU type and flags. Does anyone have an idea how I can go about debugging the issue? Or has anyone experienced this issue and found a solution? Cheers, Sebastian -- *Sebastian J. Bronner* Administrator D9T GmbH - Magirusstr. 39/1 - D-89077 Ulm Tel: +49 731 1411 696-0 - Fax: +49 731 3799-220 Geschäftsführer: Daniel Kraft Sitz und Register: Ulm, HRB 722416 Ust.IdNr: DE 260484638 http://d9t.de - D9T High Performance Hosting info(a)d9t.de

1 0

[libvirt-users] mouse pointer not in sync after live migration
by Stefan 30 Jul '12

30 Jul '12

Hi everybody, I am using Opennebula 3.6 on a Debian Squeeze system. My configuration: Node-A kvm --version QEMU emulator version 1.0 (qemu-kvm-1.0 Debian 1.0+dfsg-8~bpo60+1), Copyright (c) 2003-2008 Fabrice Bellard virsh --version 0.9.11.3 Node-B kvm --version QEMU emulator version 1.0 (qemu-kvm-1.0 Debian 1.0+dfsg-8~bpo60+1), Copyright (c) 2003-2008 Fabrice Bellard virsh --version 0.9.11.3 After live-migration, my mouse pointer is not in sync. I just write an email to the opennebula-mailinglist, but maybe this a problem from libvirt or kvm? http://lists.opennebula.org/pipermail/users-opennebula.org/2012-July/009643… thanks in advance Stefan

1 0

Re: [libvirt-users] libvirt-users Digest, Vol 31, Issue 39
by Borden Rhodes 28 Jul '12

28 Jul '12

Hey everybody, Thanks for the help. It turns out that net.ipv4.ip_forward was set to = 0 and I had to change it to = 1, which I found from the excellent http://wiki.libvirt.org/page/Guest_can_reach_host%2C_but_can%27t_reach_outs… guide. All's well in the state of Denmark and I should be able to set to figuring out how to get libvirt to play nicely with Firestarter and Network Manager. Speaking from my own experience, when people are fooling around with the documentation, they may want to consider the following things: 1) In Debian (and probably Ubuntu by association), the command to restart libvirtd is (sudo) service libvirt-bin restart. For *tards like me who get easily confused and might think that a libvirt daemon is something specific to Red Hat, etc., it might be worth making this distinction. I'll probably also whine to the Debian to make this point clearer in their own documentation since it's not the libvirt project's fault that Debian feels compelled to 'improve' things. 2) Again, a *tard like me didn't think to look for the Troubleshooting page under the General Project Documentation section of the Wiki. Granted, I realise that makes me especially *tarded because it's, like, the first item after the table of contents, but someone like me expects Troubleshooting to be its own section in the ToC and/or linked generously within the sections discussing networking concepts. Perhaps it might be more useful to start a Troubleshooting sub-subsection under SS 1.7 and link all of the network troubleshooting docs to that? 3) One of the frustrations I had was trying to figure out what the responsibilities were for each of libvirt, virt-manager and qemu/kvm, especially in the context of seeing all of the settings jumbled together in virt-manager (I'm a little bash-phobic). So I'm not sure if this is libvirt's 'problem,' but as good as Virtual Networking concepts page is, it doesn't quite explain the difference between what appears in virt-manager as 'Virtual Networks' and 'Network Interfaces' under the connection settings. I think the difference is that the former is for simple NAT routing whereas the latter is for advanced bridging, but when I was setting up my virtual machines, I didn't understand which I had to configure and whether I had to set up both. Again, thanks for the help. I really appreciate it. Borden

1 0

[libvirt-users] Can't NAT in KVM
by Borden Rhodes 28 Jul '12

28 Jul '12

Good morning, list, I've been trying to get my KVM virtual machines to connect to the Interwebs for the last 3 hours with no success. I'm running Debian Wheezy (fully patched) and building virtual machines using KVM in Virt Manager. With the default, factory settings shown in virt manager, I have a default virtual network with the device name virbr0, which starts on boot and uses the default IPv4 subnets. I also have the default lo (loopback) network interface. I have tried creating a Windows 8 Release Preview machine and a Kubuntu Live CD machine. Despite recognising the hardware and network interface, neither machine will connect to the Internet. From what I can see in both virtual machines, it seems that neither is getting a DHCP lease from virbr0. I'm not sure what's broken or where to start looking for trouble because both the logs and the documentation are fairly sparse. Other possibly relevant information: I have a simple firewall configured with Firestarter in Wheezy. I use Network Manager to manage my only wireless connection to the world. I don't want to set up any fancy bridging or override /etc/interfaces because I only use virtual machines rarely but I rely on Network Manager to set up my laptop's networks so I don't have to. Any suggestions on how I can plug the v-machines to the web? With thanks, Borden

4 6

[libvirt-users] How can I make sVirt work with LXC (libvirt-0.9.13)?
by David Kang 26 Jul '12

26 Jul '12

Hi, I've installed libvirt-0.9.13 on RHEL6.2 from the source code. I cannot make sVirt working with LXC. (sVirt works well with KVM, though.) I can start an LXC instance, but the label of the process is not right. Can someone help me? I tried to change /etc/libvirtd/lxc.conf file to explicitly enable security_driver = "selinux". But it ends up with error saying "error : virSecurityDriverLookup:74 : internal error Security driver selinux not found". SELinux is in a permissive mode but is not enforcing. The current situation is as follows: * The label of an LXC container is not properly done: $ ps auxZ | grep lxc unconfined_u:system_r:virtd_t:s0-s0:c0.c1023 root 27998 0.0 0.0 34716 1160 ? Ss 11:54 0:00 /usr/libexec/libvirt_lxc --name instance-0000001f --console 16 --security=none --handshake 19 --background --veth veth1 * This is the lable of libvirtd process: $ ps auxZ | grep libvirtd unconfined_u:system_r:virtd_t:s0-s0:c0.c1023 root 27376 0.1 0.0 919052 5324 ? Sl 11:51 0:00 libvirtd --daemon * Here is the xml file that I used to run LXC instance, which was generated by OpenStack: <domain type="lxc"> <uuid>621adc56-2103-465c-bafd-1d0b51bea3a0</uuid> <name>instance-0000001f</name> <memory>2097152</memory> <vcpu>1</vcpu> <os> <type>exe</type> <cmdline>console=ttyS0</cmdline> <init>/sbin/init</init> </os> <clock offset="utc"/> <devices> <filesystem type="mount"> <source dir="/usr/local/nova/instances/instance-0000001f/rootfs"/> <target dir="/"/> </filesystem> <interface type="bridge"> <mac address="fa:16:3e:1e:33:f4"/> <model type="virtio"/> <source bridge="br100"/> <filterref filter="nova-instance-instance-0000001f-fa163e1e33f4"> <parameter name="IP" value="10.0.0.4"/> <parameter name="DHCPSERVER" value="10.0.0.3"/> <parameter name="PROJNET" value="10.0.0.0"/> <parameter name="PROJMASK" value="255.255.0.0"/> </filterref> </interface> <console type="pty"/> </devices> </domain> * Svirt works well with KVM as is shown below: $ ps auxZ | grep qemu system_u:system_r:svirt_t:s0:c128,c132 root 22710 6.9 0.2 895040 34332 ? Sl 11:17 0:07 /usr/libexec/qemu-kvm -name instance-0000001b -S -M pc-0.14 -cpu core2duo,+lahf_lm,+dca,+pdcm,+xtpr,+cx16,+tm2,+est,+vmx,+ds_cpl,+dtes64,+pbe,+tm,+ht,+ss,+acpi,+ds -enable-kvm -m 512 -smp 1,sockets=1,cores=1,threads=1 -uuid 12716b13-001c-40de-9537-ca847aead70b -nodefconfig -nodefaults -chardev socket,id=charmonitor,path=/var/lib/libvirt/qemu/instance-0000001b.monitor,server,nowait -mon chardev=charmonitor,id=monitor,mode=control -rtc base=utc,driftfix=slew -no-kvm-pit-reinjection -kernel /usr/local/nova/instances/instance-0000001b/kernel -initrd /usr/local/nova/instances/instance-0000001b/ramdisk -append root=/dev/vda console=ttyS0 -device piix3-usb-uhci,id=usb,bus=pci.0,addr=0x1.0x2 -drive file=/usr/local/nova/instances/instance-0000001b/disk,if=none,id=drive-virtio-disk0,format=qcow2,cache=none -device virtio-blk-pci,scsi=off,bus=pci.0,addr=0x4,drive=drive-virtio-disk0,id=virtio-disk0,bootindex=1 -netdev tap,fd=17,id=hostnet0 -device virtio-net-pci,netdev=hostnet0,id=net0,mac=fa:16:3e:7a:5e:b6,bus=pci.0,addr=0x3 -chardev file,id=charserial0,path=/usr/local/nova/instances/instance-0000001b/console.log -device isa-serial,chardev=charserial0,id=serial0 -chardev pty,id=charserial1 -device isa-serial,chardev=charserial1,id=serial1 -device usb-tablet,id=input0 -vnc 127.0.0.1:0 -k en-us -vga cirrus -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x5 Thanks, David ---------------------- Dr. Dong-In "David" Kang Computer Scientist USC/ISI

1 1

[libvirt-users] libvirt shortcomings of virsh domxml-to-native and domxml-from-native
by 张光鹏 26 Jul '12

26 Jul '12

Hi,all. we use libvirt to manage some host and vms on vmware hypervisor.I have found the libvirt there are some shortcomings . Such as virsh domxml-to-native and domxml-from-native .I have reviewed the libvirt source --src/vmx/vmx.c, code as follows : libvirt0.9.13源码xml 转化成vmxvmx转化成xml /* def:os.arch -> vmx:guestOS */ if (def->os.arch == NULL || STRCASEEQ(def->os.arch, "i686")) { virBufferAddLit(&buffer, "guestOS = \"other\"\n"); } else if (STRCASEEQ(def->os.arch, "x86_64")) { virBufferAddLit(&buffer, "guestOS = \"other-64\"\n"); } else { VMX_ERROR(VIR_ERR_INTERNAL_ERROR, _("Expecting domain XML attribute 'arch' of entry 'os/type' " "to be 'i686' or 'x86_64' but found '%s'"), def->os.arch); goto cleanup; }/* vmx:guestOS -> def:os.arch */ if (virVMXGetConfigString(conf, "guestOS", &guestOS, true) < 0) { goto cleanup; } if (guestOS != NULL && virFileHasSuffix(guestOS, "-64")) { def->os.arch = strdup("x86_64"); } else { def->os.arch = strdup("i686"); } if (def->os.arch == NULL) { virReportOOMError(); goto cleanup; } Vmware support many guestOS types like "windows7srv-64 ",If I use 64-bit Windows virtual machine, other-64 this type is not correct. NIC e1000 system can not be used . So I suggest to The libvirt improve the vmware support, especially the mutual transformation of the xml and vmx format guestOS only one parameter. Thanks all. 张光鹏 tel: 13718913184 mail: zhanggp(a)neusoft.com 移动互联网事业部东软集团股份有限公司 http://www.neusoft.com --------------------------------------------------------------------------------------------------- Confidentiality Notice: The information contained in this e-mail and any accompanying attachment(s) is intended only for the use of the intended recipient and may be confidential and/or privileged of Neusoft Corporation, its subsidiaries and/or its affiliates. If any reader of this communication is not the intended recipient, unauthorized use, forwarding, printing, storing, disclosure or copying is strictly prohibited, and may be unlawful.If you have received this communication in error,please immediately notify the sender by return e-mail, and delete the original message and all copies from your system. Thank you. ---------------------------------------------------------------------------------------------------

3 2

[libvirt-users] No /dev/tty device within LXC guest?
by Yufang Zhang 26 Jul '12

26 Jul '12

Hi all, When I create a LXC guest(RHEL5.4) via libvirt, the guest could boot successfully. I could ping to guest, but can not ssh into the it. The secure log in the guest shows that this is due to lack of '/dev/tty' device: # tail /path/to/rootfs/of/container/var/log/secure Jul 27 04:25:25 xen2143v sshd[671]: Accepted password for zhangyufang from 10.0.0.1 port 34102 ssh2 Jul 27 04:25:25 xen2143v sshd[671]: pam_unix(sshd:session): session opened for user zhangyufang by (uid=0) Jul 27 04:25:25 xen2143v sshd[671]: pam_loginuid(sshd:session): set_loginuid failed *Jul 27 04:25:25 xen2143v sshd[671]: error: PAM: pam_open_session(): Cannot make/remove an entry for the specified session* *Jul 27 04:25:25 xen2143v sshd[674]: error: open /dev/tty failed - could not set controlling tty: No such file or directory* *Jul 27 04:25:25 xen2143v sshd[673]: Received disconnect from 10.0.0.1: 11: disconnected by user* The xml of the guest: # virsh --connect lxc:/// dumpxml instance-00000034 <domain type='lxc' id='30090'> <name>instance-00000034</name> <uuid>8387e0f1-6df3-41c8-82ba-cde7d2b60844</uuid> <memory>524288</memory> <currentMemory>524288</currentMemory> <vcpu>1</vcpu> <os> <type arch='x86_64'>exe</type> <init>/sbin/init</init> </os> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <emulator>/usr/libexec/libvirt_lxc</emulator> <filesystem type='mount' accessmode='passthrough'> <source dir='/var/lib/nova/instances/instance-00000034/rootfs'/> <target dir='/'/> </filesystem> <interface type='bridge'> <mac address='fa:16:3e:3c:1d:95'/> <source bridge='demonetbr0'/> <target dev='veth0'/> <filterref filter='nova-instance-instance-00000034-fa163e3c1d95'> <parameter name='DHCPSERVER' value='10.0.0.1'/> <parameter name='IP' value='10.0.0.3'/> </filterref> </interface> <console type='pty' tty='/dev/pts/0'> <source path='/dev/pts/0'/> <target type='serial' port='0'/> </console> </devices> </domain> Could anyone point me to where the problem is? Thanks. Best Regards. Yufang

1 0

[libvirt-users] VM dies with error regarding "virConnectNumOfInterfaces"
by Helga Velroyen 25 Jul '12

25 Jul '12

Hi! I ran into a problem using a USB device and was wondering if you could help me. I'm running a Win XP SP3 in a VM with libvirt on a Xubuntu 11.10. I configured it to talk to several USB devices. With one particular device (a Phonak Compilot[1], admittedly a probably rather unusual device), the VM dies with the following error in the libvirtd.log: 2012-07-20 21:04:19.591+0000: 1282: error : virConnectNumOfInterfaces:9803 : this function is not supported by the connection driver: virConnectNumOfInterfaces 2012-07-20 21:05:05.762+0000: 1276: error : qemuMonitorIO:603 : internal error End of file from monitor This happens no matter if is it the only configured USB device or if there is another USB device present in the config or not. I configured it doing the following steps: - I looked up the vendor / product ID using lsub Bus 001 Device 003: ID 180f:1002 (Interestingly, it does not show a name of the device here.) - I added it to the VM's xml configuration using virsh / edit $vmname <domain> ... <devices> ... <hostdev mode='subsystem' type='usb' managed='no'> <source> <vendor id='0x180f'/> <product id='0x1002'/> </source> </hostdev> ... </devices> </domain> - I run the VM using the virt-manager UI. Any ideas what the problem could be? Do I need to run/configure differently than how I did it? To me it sounds as if there is some implementation missing? Is there a workaround for it? I hope you can give me some clues. Thank you! Helga [1] http://www.phonak.com/com/b2c/en/products/accessories/compilot-tvlink-s/ove…

2 4

[libvirt-users] virsh iface-list error
by Arun Sharma 25 Jul '12

25 Jul '12

Hello Facing below error while trying to access the iface-list. # virsh iface-list *error: Failed to list active interfaces error: this function is not supported by the connection driver: virConnectNumOfInterfaces* root@kvm-1:/usr/local/src/libvirt# virsh list Id Name State ---------------------------------------------------- 1 vm1 running After some finding, i saw it require me to build libvirt with --with-netcf. I tried that too but still facing the same problem. Here are the steps perform on Ubuntu 1204 (64 bit) + qemu KVM. # git clone git://libvirt.org/libvirt.git # ./autogen.sh --prefix=/usr --with-qemu --with-init-script=upstart --with-yajl --with-netcf --with-network --with-init-script=upstart --with-libvirtd # make # make install # /usr/sbin/libvirtd & # /usr/sbin/dnsmasq --strict-order --bind-interfaces --pid-file=/var/run/libvirt/network/default.pid --conf-file= --except-interface lo --listen-address 192.168.122.1 --dhcp-range 192.168.122.2,192.168.122.254 --dhcp-leasefile=/var/lib/libvirt/dnsmasq/default.leases --dhcp-lease-max=253 --dhcp-no-override Pl. let me know what could be the problem? Thanks Arun

4 3