virsh, and virt-manager hang

Installing, Configuring, Troubleshooting server daemons such as Web and Mail
Post Reply
JRefL6
Posts: 3
Joined: 2013/11/06 15:22:38

virsh, and virt-manager hang

Post by JRefL6 » 2013/11/06 15:36:40

After a power hit 1 of my 2 Cent OS 5 Hosts is hanging win I try to manage the Guests through virsh.
the error message is;

Code: Select all

libvirtd: 06:57:15.286: warning : qemudStartup:1728 : Unable to create cgroup for driver: No such device or address

Code: Select all

Hanging hosts info:
uname -a
Linux MyVM-Host1.my.domain 2.6.18-371.1.2.el5.centos.plus #1 SMP Tue Oct 22 20:36:50 EDT 2013 x86_64 x86_64 x86_64 GNU/Linux

egrep  'kvm|qmeu|libvirt' installed-Apps
etherboot-zroms-kvm.x86_64            5.4.4-16.el5.centos              installed
kmod-kvm.x86_64                       83-262.el5.centos.4              installed
kvm.x86_64                            83-262.el5.centos.4              installed
kvm-qemu-img.x86_64                   83-262.el5.centos.4              installed
libvirt.i386                          0.8.2-29.el5_9.1                 installed
libvirt.x86_64                        0.8.2-29.el5_9.1                 installed
libvirt-python.x86_64                 0.8.2-29.el5_9.1                 installed

ps -ef | egrep 'kvm|qmeu|libvirt'
root      5655     1  0 06:57 ?        00:00:00 libvirtd --daemon
nobody    6012     1  0 06:57 ?        00:00:00 /usr/sbin/dnsmasq --strict-order --bind-interfaces --pid-file=/var/run/libvirt/network/default.pid --conf-file=  --listen-address 192.168.122.1 --except-interface lo --dhcp-range 192.168.122.2,192.168.122.254 --dhcp-lease-max=253 --dhcp-no-override
root      6143     1  1 06:57 ?        00:00:58 /usr/libexec/qemu-kvm -S -M rhel5.4.0 -m 4096 -smp 2,sockets=2,cores=1,threads=1 -name VM3-ASD-WEB -uuid 122b5b02-675f-424f-1be4-71efab230ea1 -monitor unix:/var/lib/libvirt/qemu/VM3-ASD-WEB.monitor,server,nowait -no-kvm-pit-reinjection -boot c -drive file=/Guests/VM3-ASD-WEB.img,if=virtio,boot=on,format=raw,cache=none -drive if=ide,media=cdrom,bus=1,unit=0,readonly=on,format=raw -net nic,macaddr=54:52:00:7d:04:5a,vlan=0,model=virtio -net tap,fd=18,vlan=0 -net nic,macaddr=54:52:00:1c:10:bb,vlan=1 -net tap,fd=19,vlan=1 -net nic,macaddr=54:52:00:3a:bd:c9,vlan=2 -net tap,fd=20,vlan=2 -serial pty -parallel none -usb -vnc 127.0.0.1:0 -k en-us -vga cirrus -balloon virtio
root      6220     1  1 06:57 ?        00:00:47 /usr/libexec/qemu-kvm -S -M rhel5.4.0 -m 6144 -smp 4,sockets=4,cores=1,threads=1 -name www-spccssl -uuid d991116b-6be3-52d1-acff-774720110114 -monitor unix:/var/lib/libvirt/qemu/www-spccssl.monitor,server,nowait -no-kvm-pit-reinjection -boot c -drive file=/Guests/VM-www.img,if=ide,bus=0,unit=0,boot=on,format=raw -drive if=ide,media=cdrom,bus=1,unit=0,readonly=on,format=raw -net nic,macaddr=54:52:00:44:7f:62,vlan=0,model=virtio -net tap,fd=18,vlan=0 -net nic,macaddr=54:52:00:44:7f:f3,vlan=1,model=virtio -net tap,fd=20,vlan=1 -net nic,macaddr=54:52:00:2f:ef:1d,vlan=2,model=virtio -net tap,fd=21,vlan=2 -net nic,macaddr=54:52:00:2f:ef:1e,vlan=3,model=virtio -net tap,fd=22,vlan=3 -serial pty -parallel none -usb -vnc 127.0.0.1:1 -k en-us -vga cirrus -balloon virtio
root      6313     1 99 06:57 ?        01:04:16 /usr/libexec/qemu-kvm -S -M rhel5.4.0 -m 6144 -smp 4,sockets=4,cores=1,threads=1 -name SPCCSQL -uuid d991116b-6be3-52d1-acff-774717846fff -monitor unix:/var/lib/libvirt/qemu/SPCCSQL.monitor,server,nowait -no-kvm-pit-reinjection -boot c -drive file=/Guests/SPCCSQL-Guest.img,if=ide,bus=0,unit=0,boot=on,format=raw -drive file=/dev/disk/by-path/pci-0000:03:00.0-scsi-0:2:1:0,if=ide,bus=0,unit=1,format=raw -drive if=ide,media=cdrom,bus=1,unit=0,readonly=on,format=raw -net nic,macaddr=54:52:00:6c:cf:e1,vlan=0 -net tap,fd=21,vlan=0 -net nic,macaddr=54:52:00:1f:ff:a1,vlan=1,model=virtio -net tap,fd=22,vlan=1 -net nic,macaddr=54:52:00:1f:ff:a2,vlan=2,model=virtio -net tap,fd=23,vlan=2 -serial pty -parallel none -usb -vnc 127.0.0.1:2 -k en-us -vga cirrus -incoming exec:cat -balloon virtio
Two of the Guests are running, the third did not start a minor problem as that is a backup system.

The system that I can use virsh on has the same versions of kernel and applications.

Code: Select all

uname -a
Linux MyVM-Host2.my.domain 2.6.18-371.1.2.el5.centos.plus #1 SMP Tue Oct 22 20:36:50 EDT 2013 x86_64 x86_64 x86_64 GNU/Linux

egrep 'kvm|qemu|libvirt' installed-Apps
etherboot-zroms-kvm.x86_64            5.4.4-16.el5.centos              installed
kmod-kvm.x86_64                       83-262.el5.centos.4              installed
kvm.x86_64                            83-262.el5.centos.4              installed
kvm-qemu-img.x86_64                   83-262.el5.centos.4              installed
libvirt.i386                          0.8.2-29.el5_9.1                 installed
libvirt.x86_64                        0.8.2-29.el5_9.1                 installed
libvirt-python.x86_64                 0.8.2-29.el5_9.1                 installed
ps -ef | egrep 'kvm|qmeu|libvirt'
root      6015     1  0 00:19 ?        00:00:00 libvirtd --daemon
nobody    6321     1  0 00:19 ?        00:00:00 /usr/sbin/dnsmasq --strict-order --bind-interfaces --pid-file=/var/run/libvirt/network/default.pid --conf-file=  --listen-address 192.168.122.1 --except-interface lo --dhcp-range 192.168.122.2,192.168.122.254 --dhcp-lease-max=253 --dhcp-no-override
root      6432     1  0 00:20 ?        00:01:47 /usr/libexec/qemu-kvm -S -M rhel5.4.0 -m 6144 -smp 2,sockets=2,cores=1,threads=1 -name www-Guest -uuid d991116b-6be3-52d1-ac4f-774720110114 -monitor unix:/var/lib/libvirt/qemu/www-Guest.monitor,server,nowait -no-kvm-pit-reinjection -boot c -drive file=/Guests/VM-www.img,if=ide,bus=0,unit=0,boot=on,format=raw -drive if=ide,media=cdrom,bus=1,unit=0,readonly=on,format=raw -net nic,macaddr=54:52:00:2f:e5:1d,vlan=0,model=virtio -net tap,fd=18,vlan=0 -net nic,macaddr=54:52:00:2f:e5:2d,vlan=1,model=virtio -net tap,fd=19,vlan=1 -net nic,macaddr=54:52:00:2f:f5:0d,vlan=2,model=virtio -net tap,fd=20,vlan=2 -serial pty -parallel none -usb -vnc 127.0.0.1:0 -k en-us -vga cirrus -balloon virtio
root      6564     1  0 00:21 ?        00:02:20 /usr/libexec/qemu-kvm -S -M rhel5.4.0 -m 6144 -smp 4,sockets=4,cores=1,threads=1 -name SPCCSQL-VMGuest -uuid d991116b-6be3-52d1-ac4f-774717846fff -monitor unix:/var/lib/libvirt/qemu/SPCCSQL-VMGuest.monitor,server,nowait -no-kvm-pit-reinjection -boot c -drive file=/Guests/SPCCSQL-Guest.img,if=ide,bus=0,unit=0,boot=on,format=raw -drive file=/dev/disk/by-id/scsi-36842b2b076e8be00156e59f7204d31e2,if=ide,bus=0,unit=1,format=raw -drive if=ide,media=cdrom,bus=1,unit=0,readonly=on,format=raw -net nic,macaddr=54:52:00:1f:ff:0f,vlan=0,model=virtio -net tap,fd=18,vlan=0 -net nic,macaddr=54:52:00:1f:ff:af,vlan=1,model=virtio -net tap,fd=20,vlan=1 -serial pty -parallel none -usb -vnc 127.0.0.1:2 -k en-us -vga cirrus -balloon virtio
root     13783     1  0 05:57 ?        00:01:08 /usr/libexec/qemu-kvm -S -M rhel5.4.0 -m 6144 -smp 2,sockets=2,cores=1,threads=1 -name ww2-Guest -uuid d991116b-6be3-52d1-bc4f-774720110114 -monitor unix:/var/lib/libvirt/qemu/ww2-Guest.monitor,server,nowait -no-kvm-pit-reinjection -boot c -drive file=/Guests/VM-ww2.img,if=ide,bus=0,unit=0,boot=on,format=raw -drive if=ide,media=cdrom,bus=1,unit=0,readonly=on,format=raw -net nic,macaddr=54:52:00:2f:e5:3e,vlan=0,model=virtio -net tap,fd=21,vlan=0 -net nic,macaddr=54:52:00:2f:f5:2e,vlan=1,model=virtio -net tap,fd=22,vlan=1 -serial pty -parallel none -usb -vnc 127.0.0.1:1 -k en-us -vga cirrus -balloon virtio
root     16849 16789  0 08:12 pts/4    00:00:00 egrep kvm|qemu|libvirt

Any suggestions?

JRefL6
Posts: 3
Joined: 2013/11/06 15:22:38

Re: virsh, and virt-manager hang

Post by JRefL6 » 2013/11/07 13:40:08

I have got the issue pined down to one of my VM's being the cause.
Disabled autostart for the suspect and virsh works as expected, do s virsh start of suspect and it is back in a hang state.

I will compare ths xml file of the problem guest with the xml file for the production version of that guest.

With luck the problem will be visible then.

JRefL6
Posts: 3
Joined: 2013/11/06 15:22:38

Re: virsh, and virt-manager hang

Post by JRefL6 » 2013/11/13 14:11:50

For What its worth.
Copying image from another system, and "virsh UNDEFINE prob-domain" followed by "virsh define domain" and "virsh start domain" seems to have cleared whatever was causing the hang when trying to start the problem domain.

I may have been able to get away with the undefine and define steps without the overhead, and time of the image copy.

Best guess is that the UPS did not last long enough to complete something before the host lost power. Time to check some other configs.

Post Reply

Return to “CentOS 5 - Server Support”