November 2014 - Users - libvirt List Archives

[libvirt-users] backup KVM qcow2 over btrfs or zfs
by Francesco Morosinotto 07 Nov '14

07 Nov '14

Hi everybody, I'm trying to implement in the non profit organization where I work a backup strategy for our VMs. At the moment I weekly backup the machine (on sunday nights) by stopping the vm, making a snapshot, exporting the xml descriptor file and syncing these files to a remote backup server. I would really like to make daily snapshot without shutting down the vms. at the moment the vms qcow2 disks are stored over an ext4 partition but I'm planning to move them over a das with a new generation filesystem. I'm trying out zfs and btrfs looking for the best one (at the moment zfs seems to be more stable and somehow easy, flexible, while btrfs seems to have better performance on my linux box). I'm choosing one of these two FS in order to use theire instant-snapshot feature; my project will be to pause the vm, dump the memory and the xml descriptor file, make a snapshot of the subvolume storing the qcow2 file and the resuming the machine back (in I hope less than 10 seconds of downtime). what do you guys suggest me? Am I going to experience bad performance having qcow2 over btrfs? I was reading something like that but the posts were really backdated... any hints is really appreciated Thank you Francesco

3 3

[libvirt-users] Any recommendation about benchmarking tools on Linux virtual machine?
by Allen Qiu 07 Nov '14

07 Nov '14

Hi all, Do you have any recommendation about benchmarking tools on Linux virtual machine for measure the potential performance hit by virtualization, especially on multi-threads computation, since I doubt if the CentOS VM can fully make use of the 32 vCPUs (out of 64 physical cores) that I allocated to it. Regards, Allen 2014-11-05 Allen Qiu

2 2

[libvirt-users] setvcpus libvirt 1.2.9
by Thomas Stein 05 Nov '14

05 Nov '14

Hello. I'm struggeling a little bit with the setvcpus command. I'm trying now to get this running for a linux VM. Here is what i have: <vcpu placement='auto' current='2'>4</vcpu> Now the VM is starting with 2 CPUs. When i try to increase the number of CPUs i get: setvcpus dev --guest 4 error: invalid argument: requested vcpu count is greater than the count of enabled vcpus in the domain: 4 > 2 Strange thing is when i set in the xml file: <vcpu placement='auto' current='4'>4</vcpu> the VM starts with 4CPUs and setting the CPUs works. I can do: setvcpus dev --guest 2 and: setvcpus dev --guest 4 Is this a bug? Should i try to upgrade to 1.2.10? There seem to be a lot of changes in it regarding setvcpus. thanks and cheers t.

3 5

[libvirt-users] setvcpus --guest on windows
by Thomas Stein 05 Nov '14

05 Nov '14

Hello. Just a quick question. Is "setvcpus $windowsVM --guest 4" supposed to work? I have qemu-guest-agent (1.6.something) running inside a windows7 vm but all i get is: virsh # setvcpus windows7 --guest 4 error: invalid argument: requested vcpu count is greater than the count of enabled vcpus in the domain: 4 > 1 The xml file looks like this: <vcpu placement='static' current='1'>4</vcpu> cheers t.

2 3

[libvirt-users] old vm cannot be migrated, new vm can be migrated
by gustavo panizzo (gfa) 05 Nov '14

05 Nov '14

scenario: i run a mix of libvirt 1.1.1 and 1.2.2 on the source hosts, 1.2.2 on the destination hosts qemu: 1.5.0+dfsg-3ubuntu5.3~cloud0 on all the hosts openstack havana and icehouse description: i have many compute nodes running libvirt+kvm, when i try to migrate guest that were created months ago it fails, if i create a vm and then migrate it works the problem is that libvirt cannot retrieve the mtu of the bridge were the interfaces are be attached, but bridge is there (up and running) $ grep qbr32740ab3-b6 /var/log/libvirt/libvirtd.log 2014-11-05 08:44:53.013+0000: 35311: debug : udevGetDeviceProperty:121 : udev reports device 'qbr32740ab3-b6' does not have property 'DRIVER' 2014-11-05 08:44:53.013+0000: 35311: debug : udevGetDeviceType:1217 : Could not determine device type for device with sysfs name 'qbr32740ab3-b6' 2014-11-05 08:44:53.013+0000: 35311: debug : udevAddOneDevice:1392 : Discarding device -1 0x7f36e80bd5c0 /sys/devices/virtual/net/qbr32740ab3-b6 2014-11-05 08:44:53.013+0000: 35311: debug : udevProcessDeviceListEntry:1414 : Failed to create node device for udev device '/sys/devices/virtual/net/qbr32740ab3-b6' 2014-11-05 08:45:08.458+0000: 35303: error : virNetClientProgramDispatchError:175 : Cannot get interface MTU on 'qbr32740ab3-b6': No such device # ip l show dev qbr32740ab3-b6 639: qbr32740ab3-b6: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP link/ether 72:d5:db:f6:b5:30 brd ff:ff:ff:ff:ff:ff # brctl show qbr32740ab3-b6 bridge name bridge id STP enabled interfaces qbr32740ab3-b6 8000.72d5dbf6b530 no qvb32740ab3-b6 tap32740ab3-b6 the difference between a new defined domain and an old one is this < <alias name='pci0'/> --- > <alias name='pci.0'/> > <timer name='hpet' present='no'/> ---> old domain does not have it i even tried patching src/util/virnetdev.c to remove the check and always return 1500 but the same thing happens i'm lost i don't know where to look at it, any clue were to look is appreciated -- 1AE0 322E B8F7 4717 BDEA BF1D 44BB 1BA7 9F6C 6333

1 1

[libvirt-users] ioremap error during bootup
by Allen Qiu 05 Nov '14

05 Nov '14

Hi all, Does anybody have an idea about this "ioremap error" during the booting stage of VM CentOS7, although it can eventually bootup successfully. Regards, Allen 2014-11-05 Allen Qiu

1 0

[libvirt-users] fstrim support in libvirt/qemu guests
by Dennis Jacobfeuerborn 04 Nov '14

04 Nov '14

Hi, I'm trying to find out how to enable fstrim support for libvirt/qemu guests. Specifically I'm trying to enable this on a CentOS 7 host (qemu 1.5). I searched the web but only found very vague information and no clear documentation on the requirements or how to modify the guest xml to enable discard/fstrim support. Can someone point me in the right direction? Regards, Dennis

4 4

[libvirt-users] The USB passthrough that refuses to die
by Mauricio Tavares 04 Nov '14

04 Nov '14

In trying to find out why my usb passthrough is not working, I decided to edit (virsh edit) the config file for the vm client, removing the following lines <hostdev mode='subsystem' type='usb' managed='yes'> <source> <vendor id='0x051d'/> <product id='0x0002'/> </source> </hostdev> checking with "virsh dumpxml" that the lines went away, and then restarting [root@vmhost ~]# virsh start nagios error: Failed to start domain nagios error: internal error Did not find USB device 51d:2 bus:2 device:3 [root@vmhost ~]# As you can see, it is still acting like the usb device is defined. What am I missing here?

1 1

[libvirt-users] KVM incremental backup using CBT
by Jd 03 Nov '14

03 Nov '14

Hi Looking in to implementing (CBT like) delta backup for KVM. The following looks promising..(last paragraph) http://wiki.qemu.org/Features/Snapshots2 * In the last para, there is a mention of copy the blocks from the disk using dirty-bitmap as reference. How to accomplish this ? block-mirror with bitmap or is there a qemu-img command ? some details would be appreciated. // backup software now reads foo.img using t0_dirty.dbmp to perform incremental backup, when finished * The backup after few runs in the backup store would be base image + bunch of delta blocks ? Will this be same as base disk and bunch of deltas ? or there is some special way to merge these ? * I am assuming this scheme (snapshot, bitmap, block merge etc.) should work with base disk in raw (non-qcow) format as well ? Right ? i.e. will it work when the storage disk is iscsi, lvm, fc ? std linux block device? * Are there libvirt-api/verbs for doing some of this or we will have to sue the qemu-monitor-command ? * What version of qemu/kvm will have the core capabilities and which min libvirt version would be sufficient ? Is there a better way to do incremental backup / CBT like backup than one mentioned here ? Thanks /Jd

4 18

[libvirt-users] quick question on schedinfo
by Yasser Shalabi 01 Nov '14

01 Nov '14

Hey everyone, Can anyone explain the scedinfo parameters? I am wondering if I can use this, (along with host side niceness scheduler tweaks) to cause vCPUs to dominate the total time of the CPUs they are pinned to. Here are the sched info parameters: Scheduler : posix cpu_shares : 0 vcpu_period : 0 vcpu_quota : 0 emulator_period: 0 emulator_quota : 0 The documentation ( http://libvirt.org/sources/virshcmdref/html/sect-schedinfo.html) is not yet filled for this. If anyone can explain these parameters, or point me to relevant code it would be helpful! Thanks Yasser

1 0