- Users - libvirt List Archives

Re: [libvirt-users] Problems moving qemu machine to libvirt
by Gunnar Thielebein 15 Jun '10

15 Jun '10

Hi Erik, thanks for the solution provided. I tried that but unfortunately it spits out this error message: > Traceback (most recent call last): > File "/usr/share/virt-manager/virtManager/engine.py", line 588, in run_domain > vm.startup() > File "/usr/share/virt-manager/virtManager/domain.py", line 150, in startup > self._backend.create() > File "/usr/lib/python2.6/dist-packages/libvirt.py", line 300, in create > if ret == -1: raise libvirtError ('virDomainCreate() failed', dom=self) > libvirtError: internal error cannot parse QEMU version number in '' Looks to me that libvirt does some version checks with the emulator-node and don't get reliable output. The wrapper script: /etc/libvirt/kvmvde: > #!/bin/bash > exec /usr/bin/kvm "$@" \ > -net nic,vlan=0,model=virtio,macaddr=52:54:61:44:21:23 \ > -net vde,vlan=0,group=libvirtd,mode=0770,sock=/var/run/vde.ctl The xml section: ... > <devices> > <emulator>/etc/libvirt/kvmvde</emulator> > <disk type='file' device='disk'> > <source file='/var/lib/libvirt/images/vlxp01-gt/tmp7Jiyzz.qcow2'/> > <target dev='hda' bus='ide'/> > </disk> ... Btw. is the bridge entry really necessary in the xml file? I wonder if this wouldn't also work, but I am not at this step, currently: => http://libvirt.org/formatdomain.html#elementsNICSSlirp "Generic ethernet connection" > ... > <devices> > <interface type='ethernet'/> > ... > <interface type='ethernet'> > <target dev='vnet7'/> > <script path='/etc/qemu-ifup-mynet'/> > </interface> > </devices> > ... Perhaps this is a place to also start the vde hub? On 10.06.2010 13:34, Erik de Castro Lopo wrote: > #!/bin/bash > exec /usr/bin/kvm $@ \ > -net nic,vlan=0,model=rtl8139,macaddr=52:54:61:44:21:23 \ > -net vde,vlan=0,group=vde2-net,mode=0770,sock=/var/run/kvm-vde-ctl Any reason no to use the virtio driver? Best, Gunnar

1 0

[libvirt-users] Functionality of virsh net-edit
by Arseny Klimovsky 14 Jun '10

14 Jun '10

Hi, I am using version libvirt on debian lenny from backports repository, version 0.7.6-1~bpo50+1, amd64. Th functionality of "virsh net-edit" command is not very clearly documented, and I think it should change a configuration of a running network. But I see it doesn't. So, I want to add new host, for example: #I have old configuration root@xxx:/etc/libvirt/qemu/networks# virsh net-edit default Network default XML configuration edited. #Here I added new line: new host definition root@xxx:/etc/libvirt/qemu/networks# cat default.xml #Here I see that new line root@xxx:/etc/libvirt/qemu/networks# virsh net-dumpxml default root@xxx:/etc/libvirt/qemu/networks# virsh net-edit default #Here I see exactly the old network definition Is this behaviour expected? If yes, how to just add new host on a running network? -- Arseny Klimovsky

2 1

[libvirt-users] Documentation Bug: struct virSchedParameter missing value field
by Guido Winkelmann 14 Jun '10

14 Jun '10

Hi, The libvirt documentation under http://libvirt.org/html/libvirt-libvirt.html#virSchedParameter fails to list the "value" member, even though that is kind of essential for using the struct. This is what the documentation says: struct virSchedParameter{ charfield[VIR_DOMAIN_SCHED_FIELD_LENGTH] field : parameter name int type : parameter type } This is how it looks in the actual header file: typedef struct _virSchedParameter virSchedParameter; struct _virSchedParameter { char field[VIR_DOMAIN_SCHED_FIELD_LENGTH]; /* parameter name */ int type; /* parameter type */ union { int i; /* data for integer case */ unsigned int ui; /* data for unsigned integer case */ long long int l; /* data for long long integer case */ unsigned long long int ul; /* data for unsigned long long integer case */ double d; /* data for double case */ char b; /* data for char case */ } value; /* parameter value */ }; This one had me scratching my head for a while, because the way it is documented, it's absolutely not clear how the struct is supposed to be used. Only after studying the virsh source code for a while did I notice that this struct apparently has more members than the documentation lets on... Guido

1 0

[libvirt-users] Missing target dev for interface
by jeremy avnet 11 Jun '10

11 Jun '10

I have over 20 guests running on a particular host and the three longest running ones are missing the network device information in their XML description: <interface type='network'> <mac address='54:52:10::f2:c5'/> <source network='local'/> </interface> Compared to the others which have: <interface type='network'> <mac address='54:52:d1:32:0a:a1'/> <source network='local'/> <target dev='vnet13'/> </interface> This is problematic as I wanted to use domifstat on the domains. Using the packet counts shown via ifconfig on the host, I was able to determine what the vnet interface should be, but passing that to domifstat doesn't work: $ virsh domifstat maki221 vnet13 Connecting to uri: qemu:///system error: Failed to get interface stats guest19 vnet13 error: invalid argument in invalid path, 'vnet13' is not a known interface How did this happen? Is there a way I can get libvirt to know about the interfaces? And the guest OS's network connections work fine. Thanks! .:. jeremy

2 2

[libvirt-users] Multipath pools - practical for use?
by Justin Clift 09 Jun '10

09 Jun '10

Hi all, Writing up some Fedora documentation, and looking to figure out the best way of mapping multipath network(!) LUNs to pools in libvirt. ie Infiniband SRP LUNs, but would probably apply to Fibre Channel equally as well. There are two approaches I can think of easily: a) Large LUNs (ie TB+) that are mapped to a host server as disk, with each LUN being configured as an LVM pool. This lets the host slice the pool up into volumes with the libvirt volume creation commands. b) Pre-purposed LUNs mapped to a host server as disk, but passed through directly to their respective guests. ie Something like Oracle, with 3 SAN LUNs mapped to guest FOO for its database data In a non-libvirt world I'd look at configuring the LUNs on the host with persistent name using multipath, then present them appropriately to whatever needs them from there. However, libvirt has a pool type called "multipath". Looking at the docs online, it doesn't appear to be all that fully featured at this point. Seems more of a brief mention and footnote, so I'm wondering whether to use it or skip it for now in the Fedora docs. Anyone have thoughts/suggestions/etc? :) Regards and best wishes, Justin Clift -- Salasaga - Open Source eLearning IDE http://www.salasaga.org

2 1

[libvirt-users] about virDomainSave and virDomainRestore
by Camilo Aguilar 09 Jun '10

09 Jun '10

Hi there, There is some special reason to require a virConnectPtr param in virDomainRestore ? Regards, Camilo

1 1

[libvirt-users] unknown procedure: 159
by DevOs Virt 09 Jun '10

09 Jun '10

Hi, I started using libvirt few days back...while using virDomainMemoryStats API, got "unknown procedure: 159" runtime error code is ------------------------------------------------------------------------------------------------------------------------------------- virDomainMemoryStatPtr statPtr; statPtr = malloc(VIR_DOMAIN_MEMORY_STAT_NR* sizeof (virDomainMemoryStatStruct)); unsigned int flags=0; int Ret; Ret = virDomainMemoryStats(domPtr,statPtr, VIR_DOMAIN_MEMORY_STAT_NR,flags); ------------------------------------------------------------------------------------------------------------------------------------- Machine details: ------------------------------------------------------------------------------------------------------------------------------------ [root@wlng-ppcore-nd3 hellolibvirt]# uname -a Linux wlng-ppcore-nd3 2.6.18-194.el5 #1 SMP Tue Mar 16 21:52:39 EDT 2010 x86_64 x86_64 x86_64 GNU/Linux [root@wlng-ppcore-nd3 hellolibvirt]# cat /etc/redhat-release Red Hat Enterprise Linux Server release 5.5 (Tikanga) ------------------------------------------------------------------------------------------------------------------------------------ virsh details: [root@wlng-ppcore-nd3 hellolibvirt]# virsh version Compiled against library: libvir 0.8.1 Using library: libvir 0.8.1 Using API: QEMU 0.8.1 Running hypervisor: QEMU 0.9.1 ------------------------------------------------------------------------------------------------------------------------------------ There are two libvirtd installations here.. ------------------------------------------------------------------------------------------------------------------------------------ [root@wlng-ppcore-nd3 sbin]# pwd /usr/sbin [root@wlng-ppcore-nd3 sbin]# ls -l libvirtd -rwxr-xr-x 1 root root 380064 Jun 10 02:33 libvirtd [root@wlng-ppcore-nd3 sbin]# ./libvirtd --version ./libvirtd (libvirt) 0.6.3 ------------------------------------------------------------------------------------------------------------------------------------ [root@wlng-ppcore-nd3 sbin]# pwd /usr/local/sbin [root@wlng-ppcore-nd3 sbin]# ls -l total 3084 -rwxr-xr-x 1 root root 3152977 Jun 10 01:25 libvirtd [root@wlng-ppcore-nd3 sbin]# ./libvirtd --version ./libvirtd (libvirt) 0.8.1 ------------------------------------------------------------------------------------------------------------------------------------ found a link which mentions libvirtd can be old hence this error is possible... not sure how to configure the latest libvirtd? How to solve "unknown procedure: 159" runtime error? Thanks, osvdev

1 0

Re: [libvirt-users] upgrading from 0.7.2 to 0.8.1
by Todd, John A 08 Jun '10

08 Jun '10

I'm still interested in knowing what vbox:///system is, but I figured out that I have manually specify -with-xen when building to get Xen support. If after I get the Xen development package and rebuild it still doesn't work, I'll drop a line here. - JT From: Todd, John A Sent: Tuesday, June 08, 2010 1:39 PM To: 'libvirt-users(a)redhat.com' Subject: upgrading from 0.7.2 to 0.8.1 Hi, I had 0.7.2 running as an out-of-the-box version of libvirtd I ran xend and it could connect to xen:/// (with virsh) I'm trying to build a eucalyptus cloud and its not working, saying 'cannot connect to xen:///' and I don't know why. So I figure I'd get the source and try and figure it out....and upgrade while I was at it. I got 0.8.1 from source, built it and installed it. First of all, it installed everything using /usr/local as its root, is this configurable? I mean doing that it left my 0.7.2 still there and I had to do some magic to replace it with the new stuff. Now it connects to vbox:///system fine, but can't see xen:/// What's vbox:///system anyway? Could it be my xend is talking to old binaries? (though I don't know what they would be exactly) Is there some configuration (beyond the certificate stuff which I did to get libvirt actually up and running) with 0.8.1 (or in general as configuration for 0.7.2 was done behind the scene as part of OS install) That I need to do so it can connect to xen:/// again? I'm not subscribed to the list, please email responses directly to me at todjo06(a)ca.com Thanks. - Q

1 0

[libvirt-users] upgrading from 0.7.2 to 0.8.1
by Todd, John A 08 Jun '10

08 Jun '10

Hi, I had 0.7.2 running as an out-of-the-box version of libvirtd I ran xend and it could connect to xen:/// (with virsh) I'm trying to build a eucalyptus cloud and its not working, saying 'cannot connect to xen:///' and I don't know why. So I figure I'd get the source and try and figure it out....and upgrade while I was at it. I got 0.8.1 from source, built it and installed it. First of all, it installed everything using /usr/local as its root, is this configurable? I mean doing that it left my 0.7.2 still there and I had to do some magic to replace it with the new stuff. Now it connects to vbox:///system fine, but can't see xen:/// What's vbox:///system anyway? Could it be my xend is talking to old binaries? (though I don't know what they would be exactly) Is there some configuration (beyond the certificate stuff which I did to get libvirt actually up and running) with 0.8.1 (or in general as configuration for 0.7.2 was done behind the scene as part of OS install) That I need to do so it can connect to xen:/// again? I'm not subscribed to the list, please email responses directly to me at todjo06(a)ca.com Thanks. - Q

1 0

[libvirt-users] Best practices on backingup/restore VMs
by Arthur Pemberton 08 Jun '10

08 Jun '10

I just moved from F11 to F13 and I'm find no information on the best way to restore VMs. I had just copied /var/lib/libvirt (with the VMs off) as part of my backup (I already backup all of /etc). But virt-manager doesn't allow creations of VMs without installing. Also, as a side note, the search feature on the website doesn't seem to allow me to search using documentation only. -- Fedora 13 (www.pembo13.com)

2 2