I don't really know where to start, as I'm not very familiar with Linux at all.
We have a LTO6 tape system bought in 2014 from Cache-A running CentOS. It came pre-installed and configured, ready to use, with both 1Gb ethernet and 10Gb ethernet adapters. End of support/end of life for the system was 1st January, 2018.
Linux kernel version: 2.6.32-220.el6.x86_64
10Gb NIC = NC552SFP
NC552SFP Driver/Kernel Module: nx_nic
NX_NIC version: 4.0.556 NC552SFP Firmware: 4.0.579
The system has up until now been connected using 1Gb ethernet, and there are no problems copying data to/from the system using 1Gb.
We have recently installed a Cisco SG550XG 10Gb switch to set up a small 10Gb network between a couple of NAS boxes and some clients. I would also like to get our LTO system up and running on 10Gb.
I can mount the server's share from a 10Gb client (Windows 10), either by Samba or NFS.
I've also tested doing the same from a laptop with ZorinOS (Ubuntu), which I connected to the Cisco switch so it would use the 10Gb IP-range. I can ping the server from the clients. I've also tested using mtu size 1500 and 9000 on both server, switch and clients. I've tested doing a direct connection as well.
I can browse the share via 10Gb, but when I initiate a copy process of some files (from the share to the client), the copy process just hangs and never start. (Windows 10 says "Calculating time.."
When I take a closer look on the server, its 10Gb network link is actually lost, and in the terminal I can't even display info about the ethernet interface no more. (For example
I have to restart the server to be able to get the NIC up and running again.
As this happens with both Samba and NFS, from both Windows 10 and ZorinOS, I don't think it's related to either Samba or NFS configuration.
Any suggestions for how I can isolate and fix the problem?