I have a theory that I'd like you to test on a specific branch. It might be an Undici bug, but I'm far from being sure. You need Node >=22.19 to make it work. Switch to the branch named fix_undici_paused_parser_crash, build it and try again.

When a network share is stalled, the easiest way to get out of it is a host reboot. Or you can try to lazy umount the share and remount it first, but no guarantee.

@TeddyAstie I think there's alot of interest in the Intel platform because of the lack of ridiculous licensing for virtualization features. The problem is that you need a pretty current 6.x kernel to use them and XCP-ng/XenServer is just stuck on a really aging kernel right now. As exposed in this thread, the PCIe pass-through isn't really working properly under XCP-ng. Even if pass-through worked, that's kinda lame because the real purpose would be to chop up the GPU into SR-IOV instances so multiple workloads could have a GPU. The Proxmox community is already rockin' and rollin' with this. I watched a video a couple of days ago where a B70 was chopped up into Qty.4, 8GB GPU's via SR-IOV and passed to clients. No ridiculous mega-$$$$ licensing. It just worked.

@dinhngtu Fantastic - thank you for thinking of and including that!