I have a 16tb hdd that I purchased to use for parity. Started rebuilding 4 days ago. Today, like a dummy, I did something that paused the parity rebuild at 95%. I tried to just “resume” but won’t resume. It still says “parity is running” so I don’t think the whole operation is lost. Would love to not have to wait another 5 days on this. Any help would be super appreciated! 

Diagnostics are attached, thanks in advance for any help!!

 

scott

 

 

buttsnake-diagnostics-20230529-1144.zip

Checked the system log and this is the error that has been thrown repeatedly. 
"May 29 08:10:59 buttsnake kernel: device offline error, dev sdh, sector 29753410160 op 0x1:(WRITE) flags 0x4000 phys_seg 30 prio class 0

May 29 08:10:59 buttsnake kernel: md: disk29 write error, sector=29753410096..."

 

Then after a zillion more of those, at the very end of these errors 

"May 29 08:10:59 buttsnake kernel: md: disk29 write error, sector=29753411200

May 29 08:10:59 buttsnake kernel: md: recovery thread: exit status: -4"

I think the usb operation was interrupted (its in a usb dock) - when i installed a usb plugin and did something without thinking. 
I looked at Common Problems and errors and it says 
"ERRORS FOUNDSUGGESTED FIX

Disk parity2 has an HPA partition enabled on itIf this is your parity disk, then you must remove the HPA partition, because its presence will impact the ability (as in you may not be able to do it) rebuild a disabled drive and/or expand your array. It is not so important if this is present on a data/cache disk. See assistance on unRaids forums for help with the commands to fix this issue. Sample of forum posts This issue mainly affects hard drives that are currently installed in, or have been in a system with a Gigabyte motherboard"

 

So I probably need to do something to remove that partition without messing up the parity  operation if possible. 

Edited May 29, 20233 yr by b1lbytr0n

USB NOT recommended for array or pool disks due to unreliable connections among other reasons.

Understood, I used usb to get the rebuild started without restarting as other operations were ongoing. I fully plan to move it, and 2 new disks to server and attach to my LSI controller. 
 

 

Edited May 29, 20233 yr by b1lbytr0n

Moving from USB to SATA is another reason USB NOT recommended for array or pool disks.

So your solution is telling me that what I have done that has failed is not recommended? That became clear to me when the operation stopped. What I’m asking for is help in resuming the parity rebuild that is at 95% so that I don’t have to start over and maybe next time I can do all the things that are recommended correctly.

If the Resume button does not work then I am not sure there is a way to resume the operation.

Thanks for your input.  I assume if I just cancel the rebuild and then put the disks in the computer, not on USB, it’ll probably go much faster. That’s probably what I’ll end up doing, but I figured I would ask for help before going nuclear.

You had to stop the array in order to assign parity2, so you already interrupted whatever else was going on.

 

The dropped connection you experienced with USB could be the least of your problems.

 

USB interfaces may not always present the disk to the server in its "native" form, which is why moving it from USB to SATA may not work well anyway.

 

Some USB interfaces might not identify the disk by its serial number, so Unraid may not be able to identify the disk, especially if you change it to SATA connection.

 

Some USB interfaces might not even present the drive size correctly and other things about it that might make parity invalid after moving it even if you could complete the parity build.

Man,  you really don't like USB. I would say thanks for your info, but not helpful at all. I'll move on. 
This  usb interface clearly does identify by the serial number, and after 4 days of rebuild with 0 errors, I'd say it's pretty stable. 
This is not a huge problem. I  can start over. Just wanted to know if any you smart guys could help me avoid doing that. 

And I didn't experienace  a dropped usb connection, I experienced an interruption to the parity operation because of a usb plugin activation. 
Drive size is correct, all that. Healthy smart info reported. This is not just a dinky cable, it's a decent hard drive dock. 

Next time, don't condescend to people who ask for help. 

30 minutes ago, b1lbytr0n said:

Just wanted to know if any you smart guys could help me avoid doing that.

May 29 08:10:59 buttsnake kernel: device offline error, dev sdh, sector 29753410160 op 0x1:(WRITE) flags 0x4000 phys_seg 30 prio class 0
May 29 08:10:59 buttsnake kernel: md: disk29 write error, sector=29753410096

 

Drive dropped offline and there were read errors, you will need to start over.

Thanks much, I went through a bunch of ssh mess and tried to restart it manually, but the device changed from sdh to sdi after it dropped offline. Thanks a ton for your nice assistance.

2 hours ago, b1lbytr0n said:

really don't like USB

USB is fine for temporary connections, some implementations are better than others for Unraid purposes. I use it all the time with Unassigned Devices plugin