- Users - libvirt List Archives

[libvirt-users] virtio and encryption
by Mike Ponomarenko 13 Aug '10

13 Aug '10

Greetings, I am trying to use virtio and encryption. My setup is: host and guest: ubuntu 10.04 server (libvirt 0.7.5-5ubuntu27, qemu-kvm 0.12.3+noroms-0ubuntu9.2) I have secret defined and disk in domain xml referencing that secret. When I start using "<target dev='hda' bus='ide'/>" in HDD config, it works. When I start using "<target dev='vda' bus='virtio'/>" I get following in syslog: qemuMonitorCommandWithHandler:240 : cannot send monitor command 'cont': Unknown error 8446744073709551594 running with "<target dev='vda' bus='virtio'/>" works fine if format is raw and there is no encryption. Is storing raw images on encrypted loop device the only way to have both virtio and ecryption?

1 0

[libvirt-users] how to make consisten backups of guests without stopping them?
by Martin Kraus 12 Aug '10

12 Aug '10

Hello. I'd like to backup kvm guests running from lvm partitions on the host without as little downtime as possible, preferably without stopping them. Could libvirt help me with this? Could anybody point me to some documentation? The question is what happens to the filesystem if I snapshot it from the host. LVM is supposed to tell the filesystem to make itself consistent but I don't know how that works if the filesystem is mounted inside the guest. thanks mk

1 0

[libvirt-users] RHEL 6 does not support bonded and vlan tagged interfaces for KVM guests?
by David Knierim 12 Aug '10

12 Aug '10

I just wrote a bugzilla report against RHEL 6 beta2 (https://bugzilla.redhat.com/show_bug.cgi?id=623199) In it I documented that I cannot get my guests to be able to use our standard network configuration, which is active-standby bonded interface with tagged vlans. The response I got was this: RHEL Product and Program Management 2010-08-11 11:39:02 EDT This issue has been proposed when we are only considering blocker issues in the current Red Hat Enterprise Linux release. ** If you would still like this issue considered for the current release, ask your support representative to file as a blocker on your behalf. Otherwise ask that it be considered for the next Red Hat Enterprise Linux release. ** While I am working with Red Hat to get this included in the official release of RHEL 6, I was hoping that I could get this working somehow. It appears to me that the problem is in the libvert-controlled network connection to the guest. Can someone give me a clue on how to get this working? Thanks, David

1 0

[libvirt-users] libvirt 0.8.3 cannot connect to remote libvirtd 0.6.3
by Haidong Zhu 11 Aug '10

11 Aug '10

It seems virsh 0.8.3 cannot connect to libvirtd 0.6.3. Machine 1: iibvirtd is 0.6.3 Machine 2: virsh 0.6.1 virsh -c xen+ssh://root@192.168.1.78 list Connecting to uri: qemu:///session Id Name State ---------------------------------- 0 Domain-0 running 1 shanxi idle Machine 3: virsh 0.8.3 virsh -c xen+ssh://root@192.168.1.78 list error: server closed connection: error: failed to connect to the hypervisor

2 1

[libvirt-users] TCP reset (RST) networking issue (maybe was 'Sporadic network issues")
by Brian Fallik 11 Aug '10

11 Aug '10

Hi, I'm seeing a very similar problem to the thread "Sporadic networking issues" [1]. In our setup, some TCP operations receive unexpected RST packets when executed on the guest. I can see these RST packets via wireshark but can't explain their source. We originally uncovered this when some 'git clone' operations failed with "Broken Pipe" in the guests. I've been able to reproduce the same failures using scp and 'svn checkout'. The failure occurs 80-90% of the time and 'svn checkout' seems to be the quickest to fail. The same operations on the host are always successful. My setup is a Lucid amd64 host with two guests (Lucid and Hardy, both i386, using kvm). Networking is configured as described in "Routed network config". One anomoly I notice is that our configuration is missing the 'mode' and 'dev' attributes of the 'forward' node. Not sure if that makes a difference, especially since networking seems to mostly work. I've tried this with both virtio and e1000 drivers with no difference. I tried Jeremy's suggestion to limit the host to start a single guest but that did not seem to have an impact. One thing I wonder is if I'm resetting the libvirt correctly to make this change. I did not reboot the host, but I did shutdown the guests, shutdown the default network, verified that no kvm was running and no tap/bridge devices existed, undefined the Hardy domain, then restarted the default network and started the Lucid kvm. Does that sound correct? Anyone have suggestions for further debugging or pointer(s) to existing related bugs? My search didn't turn up anything obvious. I'm very eager to resolve this issue since it's blocking our ability to use the VMs. Thanks, brian 1 - https://www.redhat.com/archives/libvirt-users/2010-August/msg00012.html

1 0

[libvirt-users] renaming a domain?
by Igor Serebryany 10 Aug '10

10 Aug '10

Hi, I'm trying to figure out if libvirt supports a domain rename feature, like xen's "xm rename" command. I couldn't find anything in the docs suggesting such a function. Am I missing something? Thanks! --Igor

1 0

[libvirt-users] Sporadic network issues
by jeremy avnet 10 Aug '10

10 Aug '10

I'm seeing this problem with both the virtio and e1000 driver with both linux and windows guests. About 5 - 10% of the VMs I launch have network problems. They seem to always get an IP from the DHCP server (dnsmasq), but afterwards either they can not make an outside network connection or they lose connectivity at a later time. Sometimes the connectivity comes back, sometimes it doesn't. It's very sporadic and there doesn't seem to be any rhyme or reason to it. The network is configured like "Routed network config" described here: http://libvirt.org/formatnetwork.html#examplesRoute The host machines are on Ubuntu 10.04. We were using the default Lucid packages (libvirt 0.7.5 / qemu 0.12.3), but I have upgraded us to libvirt 0.8.1 and qemu 0.12.4. I believe this has helped reduce the failure rate, but it still persists. Has anyone else had this problem or is there some suggestions for things to look into? Thank you. .:. jeremy

2 1

[libvirt-users] 0.8.3 make error
by Haidong Zhu 10 Aug '10

10 Aug '10

OS: Ubuntu 9.04 Error output: CC libvirt_parthelper-parthelper.o CCLD libvirt_parthelper libvirt_parthelper-parthelper.o: In function `main': /home/yoyoadm/libvirt-0.8.3/src/storage/parthelper.c:105: undefined reference to `ped_device_get' /home/yoyoadm/libvirt-0.8.3/src/storage/parthelper.c:119: undefined reference to `ped_disk_new' /home/yoyoadm/libvirt-0.8.3/src/storage/parthelper.c:125: undefined reference to `ped_disk_next_partition' collect2: ld returned 1 exit status make[3]: *** [libvirt_parthelper] Error 1 make[3]: Leaving directory `/home/yoyoadm/libvirt-0.8.3/src' make[2]: *** [all] Error 2 make[2]: Leaving directory `/home/yoyoadm/libvirt-0.8.3/src' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/yoyoadm/libvirt-0.8.3' make: *** [all] Error 2

1 0

[libvirt-users] howto setup linux guest with 2 nics
by Jan Paulini 06 Aug '10

06 Aug '10

hi, i have the following situation. my host system is running debian lenny and kvm. i have 2 ethernet interfaces of which i like to use eth0 for administration of the server itself without traffic from any guest systems. eth1 is connected to a different switch where i have numerous vlans. i have successfully setup a guest running virt-install using a bridge interface on the host system. here the debian config for br1 which i used: auto eth1.700 iface eth1.700 inet manual auto br1 iface br1 inet manual bridge_ports eth1.700 bridge_fd 9 bridge_hello 2 bridge_maxage 12 bridge_stp off i was now looking into the xml file for the guest hoping to find a way to use 2 ethernet interfaces for the guest one on br1 and another on br2 where br2 would be configures as br1 above on the host, just in a different vlan. how can i do that ? the network section of the xml fine lookes like this: <interface type='bridge'> <mac address='54:52:00:xx:xx:xx'/> <source bridge='br1'/> <model type='virtio'/> </interface> is there possibly a way of installing a guest with virt-install to give the guest eth0 and eth1 which are in different vlans on the hosts eth1 ? thanks for your help ! Jan

1 0

[libvirt-users] How to force the use of virtio as the guest network driver
by Tony Lewis 06 Aug '10

06 Aug '10

I'm running KVM on a standard Debian Lenny host, and I seem to be hitting the problem identified here: http://serverfault.com/questions/123566/is-anyone-using-kvm-in-production My symptom is that using rsync over SSH to transfer files into the guest gets slower and slower, and then grinds to a halt, and the machine becomes unreachable. It seems that I should be using the virtio driver, but I can't seem to make it load. The KVM guest has the virtio module, (a la find /lib/modules -name \*virtio\*) but when I try to force it by setting <model type="virtio"/> in the XML configuration, it seems to make no difference. I see that as per http://libvirt.org/formatdomain.html , "qemu -net nic,model=? /dev/null" returns the following: "ne2k_isa i82551 i82557b i82559er ne2k_pci pcnet rtl8139", but I have no qemu-kvm binary. Not sure if that's a telling symptom or not. Host kernel is 2.6.26-2-openvz-amd64, KVM version is 72+dfsg-5~lenny5, libvirt version is 0.4.6-10, all of which look good enough to use virtio according to http://wiki.libvirt.org/page/Virtio . Thanks, Tony Lewis

1 0