I recently had a 2tb data drive that failed and was running in emulated mode. My current Parity drive is 2TB as well, but my new drives are 3TB.  I followed the guide here: https://lime-technology.com/wiki/index.php/The_parity_swap_procedure but every time I click the copy button the system hangs up and becomes unresponsive to any local  input or remote connectivity.   I can power cycle the system and regain access, but no data seems to have been copied.   Looking at the system log it seems to freeze up about 30 seconds after the process begins.  I thought this could possibly be a hardware issue with my SATA connection so I made sure that both drives were connected directly to the mobo and not the PCI SATA cards that I have installed. 

 

May 10 23:05:01 Tower emhttp: import 30 cache device: sdf
May 10 23:05:01 Tower emhttp: import flash device: sda
May 10 23:05:20 Tower emhttp: copy: disk4 to disk0
May 10 23:07:23 Tower kernel: sata_sil24 0000:04:00.0: IRQ status == 0xffffffff, PCI fault or device removal?

 

I am currently running UnRaid 6.3.3

IS the command line available?  IF so, type   diagnostics   on that and it should write the diagnostics file to the logs directory/folder on your flash drive.  It might have some more clues as to what the issue is. 

Unfortunately it is not.. once I start the copy process all remote and local access to the system become unavailable. However, I can power cycle and check that log before I begin the process again.  I will post back what I find. 

 

Appreciate the assistance!

 

Don't power cycle for a couple of few hours.  Wait and see if anyone else has some ideas.  (Also, things may be proceeding normally in the background.)  Is the disk activity light still on?  Do you have access to the server?  Can you hook a monitor (and keyboard) up to it?

sata_sil24 0000:04:00.0: IRQ status == 0xffffffff, PCI fault or device removal?

 

The above error suggests an IRQ problem with the SIL controller (SIL3132?), any disks there? if possible disconnected it or use it a different slot if available.

Well I have progress!  I thought I had my Old and New Parity drives on the same sata controller but it looks like I traced the wrong sata cable in the spaghetti mess.  Once I got that part figured out and I unplugged all my other drives, it wouldn't allow me to copy.. so I went back and plugged in 2 of my working data disks to the mobo sata controller and left the cache drive and other data drives unplugged.  The copy option was then available.  The copy is now at 3% complete which is 3% further than I made it before. 

 

My guess is that my cheap PCI sata cards are just that... cheap.  They have worked for the day to day operation of my NAS for over 2 years, but are probably introducing errors that I haven't noticed.  I will need to go back and research some better PCI sata cards for my build assuming the copy completes successfully.  This is my first drive failure I have had in the 5 years I have been using UnRaid.. *knock on wood* 

 

Thanks again for the help!