[SOLVED]Disks getting disconnected randomly

[marcoslongo]

Hello.

 

So for a while now, random disks in my array are getting disabled. All is working fine and then one or more disks get disabled. I turn it off, turn it on again and all disks are available, I can add them back to the array and then rebuild and keep working. 

I changed cables and power supply.  The HDDs are not new, but seem to be in good working condition based on the smart tests.  Could this be a controller issue?  Or more likely to be disk issue?  Unraid 6.8.3

 

 

mediaserver-diagnostics-20210304-2145.zip

mediaserver-syslog-20210304-2044.zip

Edited March 31, 2021 by marcoslongo
adding info

[JorgeB]

Likely the rather common Ryzen SATA controller issue, look for a BIOS update and/or update to v6.9, newer kernel can sometimes help.

[marcoslongo]

OK cool. I see the MB has 1 bios update and 6.9 is also already available - so I will try it and see if it helps. Thanks!

[marcoslongo]

Alright - so I tried upgrading to 6.9, but I get a nice kernel panic . I tried also manually copying the bz* files to the boot pendrive, same issue.

Should I start a new thread about the system not booting or I can continue on this one?

 

[JorgeB]

That is usually a flash drive problem, try redoing it, back up current flash, redo using the UBS tool, restore config folder.

[marcoslongo]

I feel so dumb now.

 

Recreated the drive and copied the config folder - works fine. I will update in a few days if the issue with the disks go away

[marcoslongo]

Unfortunately this has not resolved. Both bios updates and unraid upgrade done. Still does not work and had one disk disconnecting today.

I'm assuming this issue is only with the internal Sata ports right? If I use an external controller I should be good?  

Need to check if my Mobo boots without a video card though.

 

Anyone has any other suggestions on how to deal with this issue?

[JorgeB]

1 hour ago, marcoslongo said:

If I use an external controller I should be good?  

Yes, use a recommended model.

[marcoslongo]

Just wanted to reply here that although I did not yet bought the controller, I'm doing a daily restart of the server. The issue happens only after a certain number of hours that is > 24h. I don't know why exactly, but it has nothing to do with number of read/writes or anything like that, it simply happens randomly after around 40h of usage. 

 

I will get a controller to stop the restart, but there it is, solved for now. Thanks for the support.