i/o usage by VM with io_uring

[systemctl]

Hello there.
Today i've faced a strange issue. atop/iotop shows a high disk write rates on the node, but the commands that does it called "io_wqe_worker-1", "io_wqe_worker-0". As far as i know qemu 6+ uses io_uring by default for virtual machines. It seems working fine but how i supposed to check what exactly VM initiates all this disk writes?
Maybe i don't understand something obvious, please help me out

Total DISK READ:        96.71 K/s | Total DISK WRITE:        35.77 M/s
Current DISK READ:     143.60 K/s | Current DISK WRITE:      39.41 M/s
    PID  PRIO  USER     DISK READ  DISK WRITE  SWAPIN     IO>    COMMAND
2998504 be/4 root        0.00 B/s    7.98 M/s  0.00 %  8.84 % [io_wqe_worker-1]
3000781 be/4 root        0.00 B/s    4.73 M/s  0.00 %  3.37 % [io_wqe_worker-1]

lsof with that pid's also does nothing

All VMs with qcow2 disks, scsi

 

[systemctl]

i founded the problem VM randomly limiting I/O operations. But that doesn't look like the best solution for me. So question is still actual