Using latest Proxmox release.
Server has CPU Xeon 4 cores 3 GHz and 10 Gb RAM.
Testing from a client with i3 2.53 GHz, 4 Gb RAM, using virt-viewer on Fedora.
I have been using a Windows 10 KVM guest for a while already, always on local network. VM has 3 Gb ballooning RAM and spice-guest-tools installed.
If using default novnc VM works acceptable.
But if using SPICE VM is rather slow in general, with constant short hangups that can make VM kind of unresponsive.
If it was, say a VM installed locally through libvirt, slow VM would make sense given client's specs. But this is not the case.
Is Spice affecting client's performance, still when VM guest is not even installed locally but in a server?
Or any other ideas?
Thanks.
Server has CPU Xeon 4 cores 3 GHz and 10 Gb RAM.
Testing from a client with i3 2.53 GHz, 4 Gb RAM, using virt-viewer on Fedora.
I have been using a Windows 10 KVM guest for a while already, always on local network. VM has 3 Gb ballooning RAM and spice-guest-tools installed.
If using default novnc VM works acceptable.
But if using SPICE VM is rather slow in general, with constant short hangups that can make VM kind of unresponsive.
If it was, say a VM installed locally through libvirt, slow VM would make sense given client's specs. But this is not the case.
Is Spice affecting client's performance, still when VM guest is not even installed locally but in a server?
Or any other ideas?
Thanks.