Most recent parity check finished today with ~500 million errors. I was shocked and checked the history  to see that the prior check in January returned similar. I don't know how/why I didn't see that at back then. It would have shocked me then too since I have only ever had 0 errors.

 

The server has been running just fine as far as I can tell. In the interim, I have upgraded the OS twice, first 6.11.5 -> 6.12.5 then ->7.0.1. Prior to the January check, but after the last "0-error" one, I did a total case migration. During this, all HDDs were moved from old norco hotswap bays to direct mounting and connections in a bigger case. Also upgraded the CPU, added 2 SSDs, and replaced all the sata cables (that aren't connected via breakout cable. 

 

I'm hoping someone can tell from the diagnostics how best to proceed.

mootower-diagnostics-20250427-1707.zip

Looks like you have upgraded parity recently, did you do a parity swap?

 

This last check was correct, run another one to confirm 0 errors.

Yes I did the parity swap procedure in October. Then in November I migrated the hardware to the new case.

If I run another check and still get errors, what would the next step be? And should I run it with "Write Corrections" enabled? It is disabled for scheduled.

Reboot, run a parity check with corrections enabled, then another one without rebooting, if the second one still finds errors, post new diags.

On 4/29/2025 at 3:45 AM, JorgeB said:

Reboot, run a parity check with corrections enabled, then another one without rebooting, if the second one still finds errors, post new diags.

the webUI froze, then reloaded with no styling/css or whatever (see screenshot 1 (ss1)) and an error that the flash drive was missing. IPMI KVM showed errors (ss2). About this time, a discord notification indicated party check completed with all the errors again (ss3), though it showed 0 errors the entire duration it was running.

 

The server then became unresponsive via all methods, so I hard-reset via the chasis power button. Upon reboot, the os loaded normally and unRaid loaded normally. The parity check was listed as complete on the Main>>Array OPerations (ss4). The boot device (ss5), ssd pools (ss6), and array hdds (ss7) are all visible.

I have diagnostics from before this parity check was initiated, and one from right after his forced cold shutdown. 

 

I've uploaded all screens and diagnostic zips as 1 archive, but If it isn't acceptable I can upload the diagnostics here and put the screens as gallery link?

 

Thanks for taking a look!

unraid troubles 2025May.zip

Diags still only show the check from the 26th, run another one now.

I will try, but it froze and failed to complete the 2nd one.

that didn't last long. I was able to boot, start the array (auto-parity-check started), then enabled local syslog server to mirror to flash and write to an array folder. Then the system became entirely unresponsive (maybe 1 minute in to the parity check). Webui, IMPI KVM nothing, so I hard power off again.

 

I think I should follow user @wirenut's lead and get another flash drive... 

 

Also I might have uploaded the wrong zip, I thought I put both diag files into the archive with all the screens, but I'll attach it here again just to be sure.

mootower-diagnostics-20250501-1510.zip

If the parity check is making the server crash, there's likely a hardware issue, since you have multiple RAM sticks, try using the server with just one, if the same try with the other one, that will basically rule out bad RAM.