- Users - libvirt List Archives

[libvirt-users] Unable to start libvirt systemd service
by Sharath Kurudi 04 Feb '19

04 Feb '19

Hi all, I am using the libvirt 4.9 version. I get the following error when I try to run the command virsh list after the system comes up virsh list error: failed to connect to the hypervisor error: Failed to connect socket to '/var/run/libvirt/libvirt-sock': No such file or directory When I run the command ps -ef | grep libvirtd, to find whether the libvirtd process is running. I get the following message. ps -ef | grep libvirtd root 418 1 0 06:33 ? 00:00:00 /usr/sbin/libvirtd --daemon --config /etc/libvirt/libvirtd.conf root 420 1 0 06:33 ? 00:00:00 /usr/sbin/libvirtd --daemon --config /etc/libvirt/libvirtd.conf After sometime, I see that the virsh list command goes through and the command ps -ef | grep libvirtd, gives the following putput ps -ef | grep libvirtd root 420 1 0 06:33 ? 00:00:00 /usr/sbin/libvirtd --daemon --config /etc/libvirt/libvirtd.conf root 610 227 0 06:34 ttyS0 00:00:00 grep libvirtd root@draco:~# virsh list Id Name State -------------------- Following are the logs of journalctl -xe | grep libvirtd -- Unit libvirtd.service has begun starting up. Feb 04 06:32:01 draco libvirtd[388]: 2019-02-04 06:32:01.298+0000: 388: info : libvirt version: 4.9.0 Feb 04 06:32:01 draco libvirtd[388]: 2019-02-04 06:32:01.298+0000: 388: info : hostname: draco Feb 04 06:32:01 draco libvirtd[388]: 2019-02-04 06:32:01.298+0000: 388: warning : virGetHostnameImpl:714 : getaddrinfo failed for 'draco': Temporary failure in name resolution Feb 04 06:33:31 draco systemd[1]: libvirtd.service: Start operation timed out. Terminating. Feb 04 06:33:31 draco systemd[1]: libvirtd.service: Failed with result 'timeout'. -- Subject: Unit libvirtd.service has failed -- Unit libvirtd.service has failed. Feb 04 06:33:31 draco systemd[1]: libvirtd.service: Service hold-off time over, scheduling restart. Feb 04 06:33:31 draco systemd[1]: libvirtd.service: Scheduled restart job, restart counter is at 4. -- Automatic restarting of the unit libvirtd.service has been scheduled, as the result for -- Subject: Unit libvirtd.service has finished shutting down -- Unit libvirtd.service has finished shutting down. -- Subject: Unit libvirtd.service has begun start-up -- Unit libvirtd.service has begun starting up. Feb 04 06:33:31 draco libvirtd[418]: 2019-02-04 06:33:31.799+0000: 418: info : libvirt version: 4.9.0 Feb 04 06:33:31 draco libvirtd[418]: 2019-02-04 06:33:31.799+0000: 418: info : hostname: draco [4 draco libvirtd[418]: 2019-02-04 06:33:31.799+0000: 418: warning : virGetHostnameImpl:714 : getaddrn Feb 04 06:33:31 draco libvirtd[418]: 2019-02-04 06:33:31.799+0000: 418: warning : virGetHostnameImpl:714 : getaddrinfo failed for 'draco': Temporary failure in name resolution -- Subject: Unit libvirtd.service has finished start-up -- Unit libvirtd.service has finished starting up. Feb 04 06:34:41 draco libvirtd[420]: 482: info : libvirt version: 4.9.0 Feb 04 06:34:41 draco libvirtd[420]: 482: info : hostname: draco Feb 04 06:34:41 draco libvirtd[420]: 482: error : virGetUserID:1041 : invalid argument: Failed to parse user 'tss' Feb 04 06:34:41 draco libvirtd[420]: 482: error : virGetGroupID:1124 : invalid argument: Failed to parse group 'tss' - msg 68/68 100% It is found that the libvirt systemd service times out, and for some unkonwn reason the previous libvitrd process is not killed. I dont see any errors in /var/log/messages or /var/log/syslog. Could anyone please help? Thanks Sharath

1 0

[libvirt-users] Hook problem
by David Gilmour 30 Jan '19

30 Jan '19

I am trying to use /etc/libvirt/hooks/qemu to control the startup of several guests with interdependencies. The goal is to delay the start of guest B until the DNS server on guest A is running. To accomplish this, I wrote a qemu hook script that detects the normal startup of guest B and start a second script in the background to wait until the preconditions to start B are fulfilled, then start B using a call to the virsh command. For this strategy to work, it must handle the case where libvirt has chosen guest B as the first guest to attempt to start. (Although renaming the symlinks in /etc/libvirt/qemu/autostart to force starting the guests in a particular order might work, I do not want to rely on this undocumented behavior). In the case where libvirt happens to attempt to start guest B before it starts guest A, the hook script needs to somehow tell libvirt to skip guest B and go on to starting the next guest. Otherwise a deadlock would result as libvirt waited for B to start, but B was waiting for A to start. I have tried to handle this by returning failure from the hook script for the initial attempt to start B once the background script has been started to implement the DNS check and eventually the delayed start of B. Unfortunately, I cannot find a way to force libvirt to continue until the background script exits. No combination of background execution, nohup, disown, setsid -f, or at seems to detach the process sufficiently to "fool" libvirt into acting on the "exit 1" line in the qemu script and proceed on to start other guests. As a result, the dependency of B on A deadlocks, and neither guest ever starts. Can someone please either find an error in my approach or propose a different strategy to implement this customized dependency of the startup of one guest on another? Thanks - ====== Here is my qemu script: #!/bin/bash if [[ "$2" == 'start' ]]; then echo "$0: Starting $1..." |& logger if [[ "$1" == 'B' ]]; then # The next line is where the background script is invoked /bin/bash /usr/local/bin/startB & # These also don't work: # (/bin/bash /usr/local/bin/startB) & ; disown # setsid -f (/bin/bash /usr/local/bin/startB) & ; disown # Unfortunately, the exit in the following line doesn't force libvirt to move on to the next guest to start until the background command has itself exited exit 1; fi fi Here is the startB script, including a call to a program named in the $dnssuccess variable that does the testing of DNS availability on guest A: #!/bin/bash until $dnssuccess do echo "$0: Delaying start of guest B 10 seconds" |& logger sleep 10; done # It's now OK to start guest echo "$0: Now starting guest B" |& logger virsh start B; =====

2 1

[libvirt-users] Keyboard problems with VNC access
by Nick Howitt 30 Jan '19

30 Jan '19

Hi, I am pretty new to libvirt, but have succeeded in setting up two VM's, Windows 10 and ClearOS (a Centos derivative) and they both have the same issue. I have installed both of them with a UK English keyboard, but the host machine is remote and in the US with a US locale. When I access either through VNC (I've tried TightVNC and VNC-Viewer), a number of my keys don't map correctly. As an example the \ key comes up with a #, and the # key comes up with a 3 (as does the 3 key). Other keys are the standard US type of mapping, but the £ (above 3) comes up blank and the $ (above 4) comes up with $ which, I think is the UK setting and not the US setting. At this point I am thoroughly confused. Does anyone have any ideas? Thanks, Nick

2 2

[libvirt-users] Workaround for "internal snapshots of a VM with pflash based firmware are not supported" problem?
by Andreas Tscharner 29 Jan '19

29 Jan '19

Hello World, We were hit by the "internal snapshots of a VM with pflash based firmware are not supported" problem. Is there a workaround for it (for which revert works)? Another format perhaps? Our use case: We have an image (Win10, UEFI) with some software installed. This image is used by Jenkins to execute some tests. After the tests, the image should be reverted to the original state. Jenkins uses the libvirt plugin to connect to and start/stop and revert the virtual machines. The images are in qcow2 format. We've usually created a snapshot of the base state using "virsh snapshot-create-as" on the host system, but this is no longer possible. What is the recommended workaround? TIA and best regards Andreas -- Andreas Tscharner sternenfeuer(a)gmail.com Gordon's Law: If you think you have the solution, the question was poorly phrased.

1 0

[libvirt-users] unable to list virtualbox domain remotely
by François Marsy 25 Jan '19

25 Jan '19

Hi dear all, I have the following issue : I have a libvirtd intstalled on an ubuntu desktop. The libvirtd is listening on port 16509 (options listen_tls = 0 and listen_tcp = 1) On the desktop I have a virtualbox installed with a running VM : "styx32-dry-run3" I want to control my VM remotely by using libvirt. I performed following tests : 1)Local virsh running fine : virsh -c vbox:///session list Id Name State ---------------------------------------------------- 2 styx32-dry-run3 running 2)Using the tcp port is showing an empty list: virsh -c vbox+tcp://127.0.0.1:16509/session list Id Name State ---------------------------------------------------- 3)net-list command works virsh -c vbox+tcp://127.0.0.1:16509/session net-list Name State Autostart Persistent ---------------------------------------------------------- vboxnet0 active no autostart yes Can any of you help me please, I really don't understand why my VM is not showing up in the 2 ) list .. and I really don't know where to look at :-( Kind regards Francois

1 0

[libvirt-users] stuck in pxe loop from uefi + tianocore/ovmf
by jsl6uy js16uy 24 Jan '19

24 Jan '19

Hello all, hope all is well Was wondering if I'm doing something wrong/missing something. I have pxe+tftp+uefi working between 2 vms. However after the build is done and I reboot, I still come up to pxe b/c tianocore always starts, from what I've seen, with trying PXE first. Great for initial builds :) but I would like to it stop after. In the xml I am pointing to the disk to boot first. if I virsh destory and run qemu directly on the qcow image, the vm comes up as expected and boots into the OS I have tried removing the nvram under /var/lib/libvirt/qemu/nvram/virtmachine. But the vm boots up and starts looking for pxe again. If I let that timeout, then type exit and select continue in the tianocore boot firmware, I just loop back into trying pxe over ipv4 etc I tried <rom bar='off'/> on the e1000 and that does work to stop pxe. Then I get to the uefi shell quickly, but all I see at that point is 2 BLK maps pointing to floppy 1 and 2. I see nothing when I try to manually add a new boot option. any help/wisdom would be appreciated. libvirtd (libvirt) 4.9.0 qemu command: sudo qemu-system-x86_64 -m 2048 -L /usr/share/ovmf/x64/ -bios OVMF_CODE.fd -drive file=/images/rdhcli02,format=qcow2,cache=writeback,if=virtio -device virtio-net,netdev=net10,mac=52:54:00:11:22:33 -netdev tap,id=net10,ifname=rdhcli02,script=no,downscript=no Is there a way to address this?

1 1

[libvirt-users] Problems with incoming multicasts
by David Gra 24 Jan '19

24 Jan '19

Hello I have realized a problem with incomming multicast packages to a VM. My Setup is: 1 Host with 1 VM. On the external Network (Host: eth0), there ist a Multicast Stream. The packages are visible at the hosts eth0. The Guest is connected via macvtap to the eth0 from the Host. Host and Guest ist CentOS 7.6 in Future it should be RedHat 7.6. The guest is connected via macvtap0 to eth0. If i performe tcpdump on eth0 of the host, i can see teh packages. If i performe tcpdump on macvtap0 on the host, i can see the packages anyway. If i performe tcpdump on the guests eth0, i can't see the multicast packages. Normal unicast communications to the external systems are still working. All my network is only IPv4. Is there a limitation to multicast on libvirt? Best regards and thank's a lot for your help!

1 0

[libvirt-users] LXC guest to have/make a device node (FUSE)
by lejeczek 22 Jan '19

22 Jan '19

hi guys With lxc container under libvirt control - how can guest create device nodes? I'm specifically looking for FUSE device. Is it possible somehow to get it to the guest(passthrough?) or allow the guest to create dev node? many thanks, L.

3 2

[libvirt-users] Slow ssh
by Pascal Legrand 22 Jan '19

22 Jan '19

Hello, i'm using libvirt/kvm/qemu on a stable Debian Linux KVM 4.9.0-6-amd64 . qemu-kvm : 1:2.8+dfsg-6+deb9u5 libvirt-daemon : 3.0.0-4+deb9u3 libvirt-clients : 3.0.0-4+deb9u3 virt-manager : 1:1.4.0-5 virt-viewer : 5.0-1 virtinst : 1:1.4.0-5 My guests are made with this kind of command line : virt-install --connect=qemu:///system --name=debian --ram=2048 --disk /home/vm/debian.qcow2,size=10,format=qcow2 --network=bridge:bridge-0-1,model=virtio --network=bridge:bridge-0-2,model=virtio --network=bridge:bridge-0-3,model=virtio --accelerate --noautoconsole --vnc --vnclisten=0.0.0.0 --keymap=fr --cdrom=/home/iso/debian-6.0.1a-i386-netinst.iso --os-type=linux --os-variant=debiansqueeze --hvm Since a long time ago now (may be a Debian upgrade) i noticed that ssh session from my personal PC to guests are very slow (connection and behaviour) My ssh configuration is the same on all my real/vm servers. The problem occure only on my guest. Have you got a solution or an idea to solve this problem ? Thanks for your help -- Pascal

2 1

[libvirt-users] *-sock created as part of normal daemon startup?
by Keith Bierman 18 Jan '19

18 Jan '19

As noted in another thread, I've not had success with homebrew/Mac OSX's libvirt install. On Ubuntu and CentOS virtual machines, installing "libvirt-daemon" and friends, I end up with various "-sock" files in /var/run/libvirt. This seems sensible. After installing on MacOS X with homebrew ls /usr/local/var/run/libvirt hostdevmgr lockd qemu storage so perhaps it isn't surprising that virsh -c vbox:/// fails. Am I right in believing that the lack of -sock files means that the "brew services start libvirt" is either misconfigured or otherwise failing?

1 0