Considering Reinstall - Persistent SMB and Network Issues, Plus VM Crashes

Hi all,

I'm reaching out for advice on a long-running issue with my Unraid system, which has been running for over 4 years. I use it primarily as a NAS, along with several Docker containers and a VM running Home Assistant. My network is segmented into VLANs, and I’ve configured virtual NICs to allow Home Assistant and some Docker containers to connect to multiple VLANs. The setup also includes a Coral PCIe, an old Nvidia GPU, PCIe network cards, and various drivers.

The main issues are as follows:

SMB and Network Problems: Transfers are unbearably slow and often disconnect, with frequent errors in the logs. These issues have been discussed in this forum several times, but none of the suggested solutions have resolved the problem. I’ve even replaced the RAM, but it hasn’t helped. Some examples of the errors I encounter include:

 

Sep 17 12:03:01 Server samba-dcerpcd[2288]:   Failed to fetch domain sid for WORKGROUP  
Sep 23 16:45:18 Server winbindd[4467]:   Returning NT_STATUS_NO_SUCH_USER  

VM Crashes: My Home Assistant VM crashes intermittently, with logs often pointing to RAM-related issues:

 

2024-10-03T18:30:33.352236Z qemu-system-x86_64: ram_block_discard_range: Failed to discard range pc.ram:17a48d000 +1000 (-22)  

I’ve set up a second Unraid system, on the same network (same VLANs, etc.) but without Docker or VMs. On this new, slightly weaker machine, SMB works flawlessly. Meanwhile, the original server continues to struggle with network and SMB issues.

I’m aware that a "reinstall" on Unraid doesn’t make much sense since the system runs in RAM. However, over the years I’ve made numerous tweaks—some via the terminal and in config files (not just through the GUI)—and I’ve lost track of many of these changes. This makes me wonder if starting fresh would help, although I don’t want to lose array data or face the challenge of reconfiguring all the drivers, networks, VLANs, etc., from scratch.

What would you recommend?
 

• Should I go for a full reinstall despite knowing it might not technically make a difference, considering all the terminal/config edits I've made?
• Is it worth going deeper into diagnosing the current setup, knowing that I might have missed something in the past?
• Additionally, is there a smart way to use the second, less powerful server as a temporary solution, and then migrate everything back to the main server once it’s fixed?

I’ve attached my diagnostics file for further insight.
 

Thanks in advance for your help!

server-diagnostics-20241004-1428.zip