Improve virtio-blk device performance using iothread-vq-mapping

I'm not sure I follow.. most features are developed in-house, some are contributed by the community. in this case, @spirit wrote the patch
 
@fabian that's basically what I was wondering. I guess a follow up request is if there's any kind of public roadmap, etc, or community engagement on said roadmap?
 
there's a very high level roadmap here: https://pve.proxmox.com/wiki/Roadmap#Roadmap

smaller features are "just" tracked in the bug tracker. most community engagement happens here in the forum and in the enterprise support channels.
 
The patch doesn't work after the latest upgrade of qemu-server. Which is really unfortunate because it'd fixed a major issue I had with a program that relies on SQLite.
 
dominiaz said:
Patch upgraded to PVE 9.0 Beta 1:

https://github.com/dominiaz/iothread-vq-mapping
Click to expand...
also, while doing
patch /usr/share/perl5/PVE/QemuServer.pm < qemuserver-iothread-vq-pve9.0.patch
its throwing some honk error. And so if we add the vq parameter the VM doesnt boot.!
But,
patch /usr/share/perl5/PVE/QemuServer/Drive.pm < drive-iothread-vq-pve9.0.patch works!.

I tried with latest Proxmox 9.0.10, and qemu-kvm 10.0.2 and qemu server 9.0.22.
Please review.
 
Last edited:
You must log in or register to reply here.
Top Bottom