Search results

Upated to PVE 9.x this week. I just wanted to let the Proxmox staff know that I appreciate them fixing the mobile web interface so it works on Firefox/Android.

The fact that it works with IOMMU disabled kind of says otherwise. Plus the fact that the kernel is binding vfio to the device says it thinks the device is being passed thru. What does "lspci -vmm" say?

The new disk is likely in the same IOMMU group as something you are passing through.

To be clear, the studies @bbgeek17 refers to are about using nvme-over-tcp vs iscsi on the host. It doesn't speak to whether nvme emulation in qemu would perform better or worse than virtio-scsi.

A lot of this kind of stuff is in qemu because people want to use it to do system emulation during SW development. It doesn't mean that it is performant or that it makes sense in a production environment.

The virtio-scsi device does not emulate any real device. That's why you need a special driver for it. What benefits could you possibly obtain by making it more complex?

Proxmox does not have a "desktop". It is a server OS and is mostly managed remotely via the web browser or SSH. It also does not have DHCP enabled by default so you will need to set a static IP by editing /etc/network/interfaces and /etc/hosts. Pick an address that is outside of your server's...

Do you pass thru any PCI devices? All memory must be pinned in that case.

Glad to hear you got it fixed. You can edit the post title to mark it SOLVED for future generations.

You can always remove Network Manager. It is absolutely not required for KDE Plasma to work.

You can in fact use Network Manager to manage any non-PVE interfaces you may have. Another Ethernet or Wifi or whatever. Just put the PVE interfaces in /etc/network/interfaces and leave the others unconfigured so Network Manager can handle them.

This is not true. KDE Plasma doesn't care how the networks are managed, just that they are up when needed. I run a VM with KDE Plasma and no Network Manager at all and it works fine. Indeed, I do not get the status of the network in my systray by why is that important for a server? It isn't as...

You don't want NetworkManager to manage vmbr0 and by default it won't because it ignores any interfaces that are managed by ifupdown. You had to override that behavior by changing the config file. Change it back to "managed = false" in the ifupdown section and it will stop complaining and vmbr0...

What CPU is the host? Maybe it is too old for RHEL 10.

I think you need to give it a package name to reinstall. ETA: Just so you are aware, there are so-called "man pages" that document most commands. See "man apt" for example.

You installed a script that modifies the PVE code. It apparently needs an update for PVE9. Maybe you could do "apt install --reinstall pveproxy" and don't install your patches to get back to a working system.

Removing subscription nag from UI.. There's the problem. You likely used one of the "community scripts" to remove the subscription nag and it installed a hook to re-apply the patch on updates, But when you go between major versions the patch is wrong because the code has changed and said...

I don't think PVE copies PAM user info to the other members of the cluster. If that is true you'd have to create the user on all cluster members.

Wouldn't life be simpler if you just use Linux to run Linux things? Install Ubuntu in a VM and take the rest of the day off.

Debian 13 uses a version of systemd that requires nesting. So it is in fact needed.