Multiple "oneimage create" operations don't work in parallel -solved

I am migrating KVM VMs from Archipel orchestrator, which includes importing mutiple VM disk images into Opennebula 5.2 datastore.

I am using CLI oneimage create command.
When I execute command it returns right away, but when I do it mutiple times for mutiple images I notice that images are getting queued - is it intentional or there is some configuration piece missing in my setup?
(4 Opennebula nodes, NFS image store)

Versions of the related components and OS (frontend, hypervisors, VMs):
Opennebula 5.2
Steps to reproduce:

Try to import multiple qcow2 images via “oneimage create”. Say
oneimage create --name "os-1" --type OS --description "os1" --path=/var/tmp/vms2/vm/drives/xxx/os1.qcow2 --driver qcow2 -d imgfsds oneimage create --name "bkp1" --type DATABLOCK --description "bkp1" --path=/var/tmp/vms2/vm/drives/xxx/bkp1.qcow2 -d imgfsds --driver qcow2

Current results:
Watch staging DIR for file import progress, you will see that until 1st image file finished copying, next image file will not start.

Expected results:
Image files should be imported in parallel independently from each other, in as many threads as many oneimage create processes are executed.

I stand corrected - they do work in parallel, in terms of parallel copying to STAGING_DIR.
However it’s not all straightforward.

I’ve started test with 3 images - 40gb OS, 45Gb home , 131Gb backup DATABLOCK.
The progress was following:
All 3 images were copying into STAGING_DIR, when smallest - OS 40gb was finished it got enabled and moved from staging to actual imageds (staging and imageds are on the same NFS share),
home 45gb and backup 131gb continued, when home image 45gb finished copying it remained on staging and backup continued copy process, home image file is NOT moved to image ds until biggest image finished. This is confusing