Write to array over network freezes UnRaid

[Zeruel]

Hello, 

I hope somebody here can help me, because I'm at my wits' end.

After moving my Unraid server to new hardware I can no longer reliably write to the array. The transfer starts, then slows within seconds while the whole system freezes (including GUI) for several minutes until the transfer fails with a network error, saying that the location can no longer be found. After a while the system will return to normal.
Parity is fine, FS seems fine and I can read anything I want from every share and disk as long as I don't write to it.
What's even "worse" is that sometimes transfers go through without a hitch. At the same time other files will cause the freeze immediately no matter which share or drive I use. (I even tried renaming them but the transfer always fails.) File sizes are usually several GB so it's not a problem of sustained writes.

I believe it started after I set up a Linux Mint VM although that could be a coincidence. I was in the process of moving new files to the server without any problems for an hour or so, when I began setting up the VM. Then the transfer failed for the first time. Later the VM was also unable to write to its own disk when trying to download updates etc. I have since deactivated VMs completely and deleted the VM along with its disk but to no avail.
Internal file transfers all work fine either through the GUI or binhex-krusader docker.
Bypassing cache has no effect.

 

My hardware:
8 HDD (one parity) with 6 coming from the old server and two new additions that cleared fine. One of the new ones has been encrypted
2 cache SSDs

Mobo is an Asrock N100M

1 (cheap) Chinese nvme 6x SATA adapter with ASM1166 chip
1 old PCIe to 4x SATA card. (Not sure what chip that is but it has been working for years in the old setup without a hitch
1 32GB RAM stick (max for the mobo) @3200 (haven't run memtest yet but the system is "stable" otherwise)

The six existing drives go to them nvme, the SSDs directly to the Mobo and the two "new" (also used) HDD go to the Pcie-Sata card.

 

Transfers and GUI access are from a W11 24H2 (insider build) desktop. 

 

Any help would be greatly appreciated as I have no clue what else I could try or what might be causing this.

server-diagnostics-20240626-1635.zip

Edited June 26 by Zeruel
typo

[JorgeB]

Lots of these:

 

Jun 26 16:01:57 Server smbd[18881]: [2024/06/26 16:01:57.843658,  0] ../../source3/smbd/smb2_service.c:772(make_connection_snum)
Jun 26 16:01:57 Server smbd[18881]:   make_connection_snum: canonicalize_connect_path failed for service Server, path /mnt/user

 

Do you have something trying to connect to /mnt/user ? Like a root share or similar.

[Zeruel]

Thank you for taking an interest!

 

No, I don't think I do. Only regular disk shares.

What would a root share look like or be used for?

[JorgeB]

It allows accessing all the Unraid shares under a single network share.

 

I do see this under SMB extras:

 

[Server]
path = /mnt/user
comment =
browseable = yes
valid users = nemo
write list = nemo
vfs objects =

 

That is not stock config, remove it and reboot, then retest.

 

 

[Zeruel]

Done and done. 
It killed all my Windows network shares.
I can still find the server on my network and see some shares but others are hidden. (Like the one for isos for example). Also can't connect to one because it rejects my username/password but I know it's correct.

This certainly shook something loose but now other things are broken. 

Oh, I forgot: That explanation reminded me. I might have set something up like that to easier access all shares on the network. (It's been a while.) But in that case that has been there for years and never caused trouble.

Second edit: File transfers still don't work. I see the same weird "freeze" and then it fails.

 

Third edit: Fixed the access problem to the shares that were acting up. Some of them lost the SMB config about which user has which access rights. Easy to restore once I found it.

Edited June 26 by Zeruel
added info

[JorgeB]

If you still see the freeze issue and it started happening after the hardware change, it suggests the problem may be hardware related, but you can try redoing the flash drive, backup the current one first and then redo it and just restore the bare minimum, like the key, super.dat and the pools folder for the assignments, also copy the docker user templates folder, if the freeze issue is gone you can then reconfigure the server or try restoring a few config files at a time from the backup to see if you can find the culprit, if it's still there it's most likely hardware related.

[Zeruel]

I was hoping to avoid that but it makes sense.
I'll give it a try.
Thank you for your help!