- Users - libvirt List Archives

Re: [libvirt-users] Attaching disks with external snapshots
by Erlon Cruz 25 Aug '16

25 Aug '16

Hi Han, Any update on this? Erlon On Wed, Aug 17, 2016 at 9:58 AM, Erlon Cruz <sombrafam(a)gmail.com> wrote: > Han, > > Did the same test again with libvirt debug enabled. This time I added the > disk[1] without snaps (OK), and then added the disk[2] with snapshots. The > libvirt log[3] (line 320) says: > > *Could not open > '/opt/stack/data/nova/mnt/a3b4c6ddd9bf82edd4f726872be58d05/volume-a75b42e8-8175-4516-b724-43ee6f3d2520': > Permission denied\r\n", "id": "libvirt-33"}]* > > But this file has 777 permissions. I can access it with any user I > tried[4]. Tough its mounted over NFS, the permissions seems to be right. > > Erlon > > [1] http://paste.openstack.org/show/556055/ > [2] http://paste.openstack.org/show/556056/ > [3] http://pastebin.com/CCFcKq5b > [4] http://paste.openstack.org/show/559018/ > > > On Tue, Aug 16, 2016 at 2:41 PM, Erlon Cruz <sombrafam(a)gmail.com> wrote: > >> Hi Han, >> >> Thanks for helping. What is the XML before the snapshot is created? Can >> you post the XML files you used too? The files are in the links pointed by >> the numbers [1], [2] etc. >> >> The versions are: >> qemu-system-x86_64 --version >> QEMU emulator version 2.5.0 (Debian 1:2.5+dfsg-5ubuntu10.4), Copyright >> (c) 2003-2008 Fabrice Bellard >> >> libvirtd --version >> libvirtd (libvirt) 1.3.1 >> >> >> >> Erlon >> >> On Mon, Aug 15, 2016 at 12:56 AM, Han Han <hhan(a)redhat.com> wrote: >> >>> Hi，Erlon >>> I was not able to reproduce your problem. It is ok to attach disk with >>> external snapshot. >>> I did it with following steps: >>> # qemu-img create /tmp/vdb 100M >>> Formatting '/tmp/vdb', fmt=raw size=104857600 >>> # virsh start snap >>> Domain snap started >>> # virsh snapshot-create-as snap s1 --disk-only --diskspec >>> vdb,file=/tmp/vdb.s1 --quiesce >>> Domain snapshot s1 created >>> # virsh dumpxml snap|awk '/<disk/,/<\/disk/' >>> <disk type='file' device='disk'> >>> <driver name='qemu' type='qcow2' cache='none'/> >>> <source file='/var/lib/libvirt/images/snap.s1'/> >>> <backingStore type='file' index='1'> >>> <format type='qcow2'/> >>> <source file='/var/lib/libvirt/images/snap.qcow2'/> >>> <backingStore/> >>> </backingStore> >>> <target dev='vda' bus='virtio'/> >>> <alias name='virtio-disk0'/> >>> <address type='pci' domain='0x0000' bus='0x00' slot='0x07' >>> function='0x0'/> >>> </disk> >>> <disk type='file' device='disk'> >>> <driver name='qemu' type='qcow2'/> >>> <source file='/tmp/vdb.s1'/> >>> <backingStore type='file' index='1'> >>> <format type='raw'/> >>> <source file='/tmp/vdb'/> >>> <backingStore/> >>> </backingStore> >>> <target dev='vdb' bus='virtio'/> >>> <alias name='virtio-disk1'/> >>> <address type='pci' domain='0x0000' bus='0x00' slot='0x0c' >>> function='0x0'/> >>> </disk> >>> >>> Save the dumped vdb's xml to /tmp/vdb.xml >>> # virsh detach-device snap /tmp/vdb.xml >>> Device detached successfully >>> # virsh attach-device snap /tmp/vdb.xml >>> Device attached successfully >>> # virsh dumpxml snap|awk '/<disk/,/<\/disk/' >>> <disk type='file' device='disk'> >>> <driver name='qemu' type='qcow2' cache='none'/> >>> <source file='/var/lib/libvirt/images/snap.s1'/> >>> <backingStore type='file' index='1'> >>> <format type='qcow2'/> >>> <source file='/var/lib/libvirt/images/snap.qcow2'/> >>> <backingStore/> >>> </backingStore> >>> <target dev='vda' bus='virtio'/> >>> <alias name='virtio-disk0'/> >>> <address type='pci' domain='0x0000' bus='0x00' slot='0x07' >>> function='0x0'/> >>> </disk> >>> <disk type='file' device='disk'> >>> <driver name='qemu' type='qcow2'/> >>> <source file='/tmp/vdb.s1'/> >>> <backingStore type='file' index='1'> >>> <format type='raw'/> >>> <source file='/tmp/vdb'/> >>> <backingStore/> >>> </backingStore> >>> <target dev='vdb' bus='virtio'/> >>> <alias name='virtio-disk1'/> >>> <address type='pci' domain='0x0000' bus='0x00' slot='0x0c' >>> function='0x0'/> >>> </disk> >>> >>> So I managed to attach disks with snapshot. >>> I want to know what is your libvirt and qemu-kvm version? >>> And can you provide your files: /tmp/disk.xml[1], >>> /tmp/snap-from-disk.xml[2], /tmp/disk-with-snap.xml[4], >>> /tmp/disk-with-snap.xml . I need more details. >>> >>> Han Han >>> >>> ----- Original Message ----- >>> From: "Erlon Cruz" <sombrafam(a)gmail.com> >>> To: libvirt-users(a)redhat.com >>> Sent: Friday, August 12, 2016 10:23:24 AM >>> Subject: [libvirt-users] Attaching disks with external snapshots >>> >>> Hi folks, >>> >>> I'm trying to to attach a disk to an instance using libvirt. The problem >>> is, this disk has external snapshots. The process tried was: >>> >>> 1 - Attach a disk in the domain: >>> virsh# attach-device instance-00000006 /tmp/disk.xml[1] --live >>> >>> 2 - Snapshot the disk[2]: >>> virsh# snapshot-create instance-00000006 --quiesce --xmlfile >>> /tmp/snap-from-disk.xml[2] --disk-only >>> >>> 3 - Dump the domain XML and create a new disk file from it: >>> virsh# dumpxml instance-00000006 >>> ... [3] >>> >>> 4 - Dettach the device and re-add it using the new disk file: >>> virsh# detach-device instance-00000006 /tmp/disk-with-snap.xml[4] >>> Device detached successfully >>> virsh# attach-device instance-00000006 /tmp/disk-with-snap.xml >>> error: Failed to attach device from /tmp/disk-with-snap.xml >>> error: internal error: unable to execute QEMU command 'device_add': >>> Property 'virtio-blk-device.drive' can't find value 'drive-virtio-disk1' >>> >>> Question, is this operation supported? If yes, how is the correct >>> procedure? >>> >>> Erlon >>> >>> >>> [1] http://paste.openstack.org/show/556055/ >>> [2] http://paste.openstack.org/show/556056/ >>> [3] http://paste.openstack.org/show/556063/ >>> [4] http://paste.openstack.org/show/556064/ >>> >>> _______________________________________________ >>> libvirt-users mailing list >>> libvirt-users(a)redhat.com >>> https://www.redhat.com/mailman/listinfo/libvirt-users >>> >> >> >

2 3

[libvirt-users] qemu+ssh authentication
by Shahar Havivi 22 Aug '16

22 Aug '16

Hi, I try to login via the libvirt.openauth() API with qemu+tcp protocol and prompt for password, I am setting the username, password and the callback methdod with the following keys: auth = [[libvirt.VIR_CRED_AUTHNAME, libvirt.VIR_CRED_PASSPHRASE], request_cred, None] When I call openauth() method: conn = libvirt.openAuth('qemu+ssh://...', auth, 0) I am getting prompt for entriyng password. When using qemu+tcp it works fine. Thank you, Shahar.

2 1

[libvirt-users] libvirt-lxc capabilities mknod
by jsl6uy js16uy 19 Aug '16

19 Aug '16

Hello all, hope all is well Issue: Any way to give granular mknod capabilities to a container? Only allow creation of specific device? bit of background Have a laptop running arch and libvirt loading an arch lxc container created from lxc-create Overall container is up and running, I use it for vpn connections Initially it would not setup of the tun device. Previously using just the lxc tool set, I can edit the lxc.conf config file for the container and allow device creation of just the tun device. In libvirt I can add capabilities for mknod, but seems to be blanket for any device creation within the container? Is this correct? Thanks and Regards

2 2

[libvirt-users] Facing problems while connecting virsh to XenServer 6.5 to import VMs to RHEV / oVirt
by Anantha Raghava 18 Aug '16

18 Aug '16

Hi, I am trying to import couple of important Virtual Machines from XenServer 6.5 (from downloaded from xenserver.org) into RHEV / oVirt environment. When I used the RHEV / oVirt Import Wizard, I was facing the issues. In order to check further, I attempted to use virsh -c ssh://root@<xenhost> command to connect and check but received the same set of errors as in RHEV / oVirt Wizard. I am able to connect to xen server 6.5 using ssh command without password and even "sudo -u vdsm ssh root@<xenhost>" also works without any issues. To dig further I, on xen server, disabled the firewall and set SELinux to disabled and I am finding that libvirt is unable to connect to hypervisor. I have attached the libvirt client logs for your reference. Can someone help me to debug this and import the VMs from Xen Server 6.5 to RHEV / oVirt? We cannot recreate these Virtual Machines as these are critical ones and rebuilding will be a herculean task. Thanks in advance. -- Thanks & Regards, Anantha Raghava Do not print this e-mail unless required. Save Paper & trees.

2 2

[libvirt-users] Python API does not return all information available
by David Ashley 16 Aug '16

16 Aug '16

All - The Python virDomain method info() does not return all the entries contained in the C struct virNodeInfo. While this is somewhat documented the entries that are returned are unlabeled and the documentation does not specify which VirNodeInfo entries they correspond to. Does anyone have more information they can pass on? Thanks, Davis Ashley

2 1

[libvirt-users] Definition /dev/shm prevent LXC container from start: "Failed to access '(null)': Bad address"
by mxs kolo 15 Aug '16

15 Aug '16

Hi all We use libvirt 1.2.x and 1.3.x with next filesystem definition in config: <filesystem type='ram' accessmode='passthrough'> <source usage='524288' units='KiB'/> <target dir='/dev/shm'/> </filesystem> But libvirt 2.1.0 with same config report error: Failed to access '(null)': Bad address Debug level 1 show: 2016-08-11 10:14:56.661+0000: 1: debug : lxcContainerMountAllFS:1618 : Mounting all non-root filesystems 2016-08-11 10:14:56.661+0000: 1: debug : lxcContainerMountAllFS:1625 : Mounting '(null)' -> '/dev/shm' 2016-08-11 10:14:56.661+0000: 1: error : lxcContainerResolveSymlinks:629 : Failed to access '(null)': Bad address 2016-08-11 10:14:56.661+0000: 1: debug : virFileClose:102 : Closed fd 15 2016-08-11 10:14:56.661+0000: 1: debug : virFileClose:102 : Closed fd 10 2016-08-11 10:14:56.661+0000: 1: debug : virFileClose:102 : Closed fd 12 2016-08-11 10:14:56.661+0000: 1: debug : lxcContainerChild:2271 : Tearing down container Failure in libvirt_lxc startup: Failed to access '(null)': Bad address Someone use libvirt 2.x with /dev/shm device ? Probably need change config and add some definitions to /dev/shm ? b.r. Maxim Kozin

2 3

[libvirt-users] Guest Agent and API Relationship
by David Ashley 14 Aug '16

14 Aug '16

All - I have looked everywhere I can think of (except the C code) to see if I can find the relationship of an API to the guest agent. What I want to know is whether or not a given API has a dependency on the guest agent being available. I know that some of the APIs now have such a dependency but there does not seem to be an indicator in the documentation for when this is the case. Nor is there any comprehensive list of APIs that have a dependency on the guest agent. Any help would be useful. Thanks, W. David Ashley

2 1

[libvirt-users] Attaching disks with external snapshots
by Erlon Cruz 13 Aug '16

13 Aug '16

Hi folks, I'm trying to to attach a disk to an instance using libvirt. The problem is, this disk has external snapshots. The process tried was: 1 - Attach a disk in the domain: * virsh# attach-device instance-00000006 /tmp/disk.xml[1] --live* 2 - Snapshot the disk[2]: * virsh# snapshot-create instance-00000006 --quiesce --xmlfile /tmp/snap-from-disk.xml[2] --disk-only* 3 - Dump the domain XML and create a new disk file from it: *virsh# dumpxml instance-00000006* * ... [3]* 4 - Dettach the device and re-add it using the new disk file: * virsh# detach-device instance-00000006 /tmp/disk-with-snap.xml[4]* * Device detached successfully* * virsh# attach-device instance-00000006 /tmp/disk-with-snap.xml * * error: Failed to attach device from /tmp/disk-with-snap.xml* * error: internal error: unable to execute QEMU command 'device_add': Property 'virtio-blk-device.drive' can't find value 'drive-virtio-disk1'* Question, is this operation supported? If yes, how is the correct procedure? Erlon [1] http://paste.openstack.org/show/556055/ [2] http://paste.openstack.org/show/556056/ [3] http://paste.openstack.org/show/556063/ [4] http://paste.openstack.org/show/556064/

1 0

[libvirt-users] QEMU IMG vs Libvirt block commit
by Erlon Cruz 12 Aug '16

12 Aug '16

Hi folks, I'm having a issue in the standard NFS driver on OpenStack, that uses qemu-img and libvirt to create snapshots of volumes. It uses qemu-img in the Controller to manage the snapshots when the volume is not attached (offline) or calls the Compute (which calls libvirt) to manage snapshots if the volume is attached (online). When I try to create/delete snapshots from a snapshot chain. There are 3 situations: 1 - I create/delete the snapshots in Cinder (it uses qemu-img). It goes OK! I can delete and I can create snapshots. 2 - I create/delete the snapshots, in online mode (it uses libvirt). It goes Ok, as weel. 3 - I create the snapshots in Cinder (offline) and delete then in online mode (using libvirt), then it fails with this message[1]: libvirtError: invalid argument: could not find image 'volume-13cb83a2-880f-40e8-b60e-7e805eed76f9.d024731c-bdc3-4943-91c0-215a93ee2cf4' in chain for '/opt/stack/data/nova/mnt/a3b4c6ddd9bf82edd4f726872be58d05/volume-13cb83a2-880f-40e8-b60e-7e805eed76f9' But, in this folder, there the backing files are there[2] and they are chained as think it suppose to be[3]. The version for the 2 hosts in tests are: Controller/Cinder node: qemu-img version 2.0.0 && libvirt 1.2.2-0ubuntu13 Compute node: qemu-img version 2.5.0 && libvirt-1.3.1-1ubuntu10 Is there any compatibility problem between libvirt and qemu-img snapshots? Have you guys found any problem like that? Erlon ---------------------------------- [1] http://paste.openstack.org/show/543270/ [2] http://paste.openstack.org/show/543273/ [3] http://paste.openstack.org/show/543274/

1 2

[libvirt-users] Go bindings to LibVirt
by Sergey Bronnikov 11 Aug '16

11 Aug '16

Hi, everyone for a small project I need a Go bindings to LibVirt library. There is a page on the official site with a list of bindings to different languages [1], but it is lack of Go bindings. Does it mean there are no official bindings for that language? [1] https://libvirt.org/bindings.html Sergey

2 1