I upgraded to Unraid 7 from 6.(something, I always install the offered updates). Then ran parity check to prepare to replace one of the drives with a new one. The new drive was mounted as unassigned device.

 

Usually parity checks never report errors. This time it found 5 errors. After a day when it reached about 70% the time left shown switched from a few hours to 60 days, the speed was very slow and the amount of reads for the disks were displayed as 4 digits and 3 dots. I was able to hover over the read amounts and saw them change to full numbers in the display.  I have never before seen this weird behavior.

 

I tried to stop the parity check, but no response. After a while I got scared for my disks and shut down the server.

 

When I rebooted I tried another parity check. It found the 5 errors and made it all the way through. I started another parity check to make sure those 5 errors were really gone. That check found errors again, and then just froze. I was not able to stop it, restart the server, nothing. I pressed the off button once to get a controlled shut down, but nothing happened. I had to hold the off button to shut the poor machine off. I unplugged my unassigned new disk to make sure it wasn't causing the weird behavior.

 

When the server restarted, the parity check started as well, but I was able to stop it. I waited a while, and then started parity check again. It again found some errors, even though it should have fixed those in the prior check. This was very weird, and I thought maybe there is something on my old hardware (motherboard from 2012) which does not work with Unraid 7.

 

Rolled back OS to 6.something (the rollback process chose the version), and started the parity check again. It found 2 errors which it corrected. To make sure the parity is really clean now I started another parity check in non-corrective mode.

Server log says it installed the plugins etc at reboot, and had error messages complaining about running a too low  OS 6 version, and about not running OS version 7, then the clock became unsynchronized, and the system rebooted itself. It auto-started the parity check again, but with correction switched on.

 

After seeing all this stuff in the syslog I gave up, took one last diagnostics, and shut the system down. Those last diagnostics are attached.

 

If needed I can upload more diagnostics: I took several during those failed parity tests

 

Hardware:

 

ASUS P8Z68 with Intel i7 processor

6 port ASM166 SATA controller

 

Drives:

 

Parity:  2 Seagate Exos 20TB

7 Data disks:  2 Seagate Barracuda Pro 8TB
               2 Seagate Exos 8TB
               1 WD Black 8TB
               1 Seagate Ironwolf 10TB   

 

Thank you for reading this. Any suggestions on how to proceed are greatly appreciated
 

muellkippe-diagnostics-20250227-1516.zip

Edited March 5, 20251 yr by Juniper
problem solved --> topic edited to [SOLVED]

8 hours ago, Juniper said:

and the system rebooted itself.

This is almost always a hardware issue, rest of the symptoms also point to that, start by running memtest

Thank you much for your help!  memtest found an error.  I need new memory.

 

Thank you again. Will report back how server is doing with new memory.

 

Edit 03/05:

All set:  new memory installed and server successfully updated to unraid 7.01. Parity check found 0 errors, and I was able to replace my drive.

Thank you again for your help!

Edited March 5, 20251 yr by Juniper