Tag Info

New answers tagged

0

The used memory are claimed pages, not necessarily used in the VM. KSM can clear those if this is critical (I usually run it when the host hits 80% memory usage) Every virtual CPU is a process, so you get a process for every virtual core, not per VM. This is also what allows KVM to avoid gang scheduling, like you see in VMWare.


1

I'd recommend that you get the second NIC on the physical host connected. Use one of them for "Management" purposes of the host itself, and the other for "VM traffic". They'll probably uplink to the same place, but this at least gives you access to the host even if something goes sideways with your VMs. Once you've got that configured, you can easily do ...


-1

I am familier with vmware environment where each virtual server will be allocated with virtual NIC and vmware tools installed. virtual servers can be either be assigned with static IP or automatic as long as there is DHCP server in the same network. Apologies, I couldn't be specific to your question but I thought I would share my experience with vmware.


0

SOLVED! ANSWER: DO NOT OVERWORK entire week long 16-18hours a day until 4:30am and expect your brain to not be FRIED! IT WAS A STUPID WINDOWS FIREWALL ISSUE!!! YES, I "THOUGHT" I had disabled it but apparently I was too sleepy to have noticed I only disabled domain and maybe outgoing but not incoming firewall! OK, I'm going to try and get some sunlight ...


0

QEmu/KVM provides the options to use a TAP interface in your host: when your host send an Ethernet frame to this interface, it is sent to the interface of your guest; when your guest sends an Ethernet frame to its interface, it is srnt to the TAP interface in your host. You can create both VMS with TAP interfaces. Now you need to exchange the frames ...


0

You can try to use the macvtap driver. Note that this does come with some limitation - be sure to read libvirt documentation about it.


0

Nevermind. I forgot that I installed Kimchi on there to manage it. Killed the process and everything is much better. Derp...


-1

I wrote up my experience here: http://www.returnbooleantrue.com/2015/04/making-your-windows-kvm-guest-boxes-fly.html You can certainly add VirtIO after you finish installing windows.


0

General option 1 It's possible to share a physical drive, or in general, any block device between two KVM being run on the same host. Just pass device filename to -drive option. General option 2 If you need to share the drive between two VMs on two different hosts (in general), you need SAN (storage access network) in general. It could be FibreChannel ...


1

Two options the way I see it: Use NFS as per other commenters - NFS seems to be the path of least resistance juggle disk assignment and use some orchestration tool with KVM you can do things like "hot-plug" disk etc. via: $ virsh attach-disk ... then inside VM that got this disk attached mount disk and do your backup umount disk inside VM on ...


2

Your DNAT rule will route all TCP traffic, received by your KVM host, on port 25, to KVM guest, no matter the interface or destination. You should add either interface or destination IP (or both) in that DNAT: iptables -t nat -A PREROUTING -i your_pub_if -p TCP --dport 25 -j DNAT --to-destination 192.168.122.201:25 or iptables -t nat -A PREROUTING -d ...


2

Easy solution: connect with nc -U socket-file to your QEMU monitor, where socket-file is the path to monitor. How to find: ps -FA | grep qemu. You will find there at the command line. And run the command at the monitor: device_add usb-host,id=<ANY string>,hostbus=<BUS>,hostport=<PORT> this is pass-through by port number. You can pass ...


1

You need to generate a Virtual MAC address for the failover IP in your OVH manager, and then assign that MAC address to the virtual machine's NIC. OVH does not allow devices with arbitrary MAC addresses to communicate on its network. The virtual machine should use a bridged network, not the "default" NATted network (such VMs can talk to the Internet but ...


0

There are 2 sorts of VM VNC, VNC provided by hypervisor (aka. qemu). And VNC provided by the VM self. Obviously , libvirt and qemu have no idea of 'VNC provided by the VM self', the cmd 'virsh vncdisplay' actually extract vnc bind info from 'VNC provided by hypervisor', while your virt-install cmd line didn't instruct libvirt to do so. So, your case ...


1

On old Qemu/KVM versions, Qcow2 backend was very slow when not preallocated, more so if used without writeback cache enabled. See here for more information. On more recent Qemu versions, Qcow2 files are much faster, even when using no preallocation (or metadata-only preallocation). Still, LVM volumes remain faster. A note on the cache modes: writeback ...


1

If you're running your vms with a single command, for arguments you can use kvm -drive file=/path_to.qcow2,if=virtio,cache=off <...> It got me from 3MB/s to 70MB/s


0

I just did this today - pretty much as 'Fox' described, except change the pci controller entry instead of deleting it - changing 'pci-root' to 'pcie-root' Then it will complain about anything with a bus-related parameter - I think you can leave the item but remove the specification. I also had to remove USB-related entries (piix3 integrated controller ...


2

To update the accepted answer. Even in Fedora 20 with latest updates I had same issue running RHEL 5.3 guest. To make the above workaround persistent through reboot put following in /etc/sysconfig/modules/kvm.modules #!/bin/sh if [ ! -c /dev/kvm ]; then /usr/sbin/modprobe kvm_intel emulate_invalid_guest_state=0 > /dev/null 2>&1 fi Make the ...


2

First, I totally agree with @dyasny, it is hard to find a reasonable use case for 'full VM state (aka. with memory)'. But, if you really want 'virsh save vm_name memdump' without destroy the vm, you can try virsh snapshot-create-as ${domain} ${fake_snap} 'save vm while keep running' \ --no-metadata --atomic --live \ --memspec ...


1

Don't know about proxmox, it's a weird hybrid thing, but you could use qemu-nbd to achieve what you want. The trick is - you can't know when the guest might want to alter the files that are being tar-ed, and if that happens, the backup you're taking is worthless.


1

The problem was using a disk image from DreamSpark/OnTheHub. I downloaded a new image, and it worked out of the box. The old image name was en_windows_server_2012_r2_x64_dvd_2707946.iso. I "ordered" a new free Windows Server 2012 of DreamSpark. I believe redownloading using Secure Download Manager will download the old image again, but I haven't tested ...



Top 50 recent answers are included