I did not know about this forum until a couple of days ago. If anyone is interested i have an git repo which has been laying around for some time now, which makes the plotting easier (at least for me).
Dang it, i have a hard time remembering if i used the plot mover in this program or if i let the madmax plotter program do that.
Nevertheless, in my case, i created my plots on one server and moved them to another server (supermicro with jbod).
Let me explain my setup.
Plotting server:
Lenovo server with 56 virtual cores, 256 GB ram, a simple ssd running ubuntu os, x2 Samsung 980 PRO 1TB M.2 NVMe in raid 0 and a 10 gbps nic.
=> The madmax created a plot each ~20 min and to the ram drive and outputted the plots to the NVMe drive (from what i can remember).
=> Samba share was “activated” on the NVMe raid 0 drive
Farming server:
An old supermicro server with 16 virtual cores, a 10 gbps nic and windows server 2019 (other specs is not important )
=> This server had a mapped network drive to the samba share of the NVMe raid 0 drive
=> I then had a python program running (link below), which moved each plot to its actual destination drive. In my case a bunch of 12 TB samsung SAS drives
=> Unfortunately i dont remember the transfer speed over the 10 gbps link, but dang it was fast I think the bottleneck where the actual destination drives, so the transfer speed was at least 200 mb/s