- Users - libvirt List Archives

Re: [libvirt-users] Excluding RAW disk from snapshot
by Trey Dockendorf 03 Aug '12

03 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

[libvirt-users] test - ignore
by Harish Patil 03 Aug '12

03 Aug '12

test

1 0

[libvirt-users] Excluding RAW disk from snapshot
by Trey Dockendorf 02 Aug '12

02 Aug '12

I have an oVirt engine node that currently runs the system on a qcow2 based virtual disk, and an NFS export volume on RAW. I'd like to take a snapshot before a somewhat risky upgrade of oVirt and can't with the raw image attached. Looking at the documentation for the snapshot XML, I created this XML file for the snapshot <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... # virsh snapshot-create dc-vmarchitect dc-vmarchitect_ss.xml error: argument unsupported: unable to handle disk requests in snapshot This occurred on both libvirt-0.9.4 and after an upgrade to libvirt-0.9.10 Current system is CentOS 6.2 with 6.3 versions of qemu/libvirt qemu-kvm-0.12.1.2-2.295.el6_3.1.x86_64 qemu-img-0.12.1.2-2.295.el6_3.1.x86_64 libvirt-client-0.9.10-21.el6_3.3.x86_64 libvirt-python-0.9.10-21.el6_3.3.x86_64 libvirt-0.9.10-21.el6_3.3.x86_64 According to the documentation, http://libvirt.org/formatsnapshot.html, the disk features should be available since 0.9.5. Is this still missing or have I created an invalid XML? Thanks - Trey

2 1

[libvirt-users] Using virsh to load scripts for the guest machine
by Mauricio Tavares 02 Aug '12

02 Aug '12

Right on the top of http://www.centos.org/docs/5/html/5.2/Virtualization/chap-Virtualization-Ma…, it seems to imply you can load/send scripts to the vm guest using virsh. Is that possible? How and what are the limitations? Can you query the vm guest?

4 7

[libvirt-users] libvirt windows error
by zhijun liu 02 Aug '12

02 Aug '12

hi,all,I install the libvirt windows version ,now i type "virsh" in cmd env:windows7 (ip:192.168.0.15) the error is : info : libvirt version: 0.8.8 error : vboxLookupVersionInRegistry:401 : Could not parse version number from '%VER%' what's problem.thanks advance!!!

3 2

[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