FYI, a fix is still on the way and hopefully land soonish: https://lists.proxmox.com/pipermail/pve-devel/2024-July/064420.htmlUnfortunately I'm seeing this same issue still in an proxmox 8.0 -> 8.2 live migration. (Same VM as last time).
My work around is still to just stop the IO intensive daemon inside that vm during the last fase of the migration.
My original patch was rejected, because it would hurt performance too much and doing it properly without that performance penalty required upstreaming a new feature in QEMU, which took a while, and landed in QEMU 8.2: https://gitlab.com/qemu-project/qemu/-/commit/2d400d15a02dca3b7b90761b2f0bb2322e99e11a
The patch on the pve-devel list is then for making use of the feature in Proxmox VE's qemu-server management code.