So, I think I screwed myself tonight, looking for help. 6.11.5+

[trurl]

5 minutes ago, mattw said:

I now have the option to run with correction.  I assume that is the next step?

yes

 

On 12/20/2022 at 11:53 PM, trurl said:

without the -n. If it asks for it use -L. Capture output and post it.

 

[mattw]

Attached completed log.

disk check.txt tower-diagnostics-20221222-1023.zip

[JorgeB]

Now mount the disk with UD and compare the data with the emulated disk, see which one is better.

[mattw]

So, I am pulling some stuff off of the share for the physical disk.  Seems that the recoveries were different between the 2, neither really seems better just different.

[JorgeB]

You can rsync from one to the other, so that you have any different data from both in one disk.

[mattw]

So, I have what I want and need off of the emulated and real drive.  At this point, I think the emulated may be a little better.

[JorgeB]

In that case just rebuild on top of the old disk.

[mattw]

Can you give me a little more instruction... I have proven that I can really screw things up here! 

[trurl]

If you're sure you don't need anything still on the original disk now, you can rebuild the contents of the emulated disk onto that original disk by stopping the array, reassigning the disk to its slot in the array, and starting the array to being rebuild. Note that the contents of that original disk will be completely overwritten by the contents of the emulated disk.

 

Please let us know if you have questions or other problems.

[mattw]

Rebuild running, getting 120MBs... only 4.5 hours and I hope the power stays on.  I assume that if we have power drop, I can stop the action in the array operations window, power down and start again at next boot?

[trurl]

7 minutes ago, mattw said:

power down and start again at next boot?

You can but it will start over again from the beginning.

[itimpi]

39 minutes ago, mattw said:

Rebuild running, getting 120MBs... only 4.5 hours and I hope the power stays on.  I assume that if we have power drop, I can stop the action in the array operations window, power down and start again at next boot?

Do you have a UPS so that Unraid can shutdown tidily if the power goes out?   If so it might be worth installing the Parity Tuning plugin as it could resume the rebuild from the point reached at next boot as long as Unraid was able to do a tidy shutdown.   Standard Unraid always restarts unfinished array operations from the beginning.

[mattw]

Well, at my desk my UPS is very small.  If we start taking power hits I will manually down the machine.  One of the things I was prepping to do when i broke the array was to get the UPS and server connected and talking, did not get that far.  I will keep that in mind, hope I do not have to do rebuilds to often. 

 

Gotta ask, your sig indicates test VMs with Unraid versions... how on earth does that work with you array?  Can you only have one started at a time and it runs the array?

[mattw]

So, the array looks good now and I am moving things back without a cache drive.  Really want to put it back in, but I am not sure if the cache was the root of the problem or if I had bad ram.  I am leaning toward bad ram as the system is dealing well with large moves now, just slow.

[trurl]

11 minutes ago, mattw said:

the array looks good now

Run a non-correcting parity check to confirm.

[mattw]

Ok, I did not anticipate that step and had moved a bunch of stuff back to the array.  That is going to make they check look bad?  Although it was updating parity as it moved stuff in.

[itimpi]

2 hours ago, mattw said:

Ok, I did not anticipate that step and had moved a bunch of stuff back to the array.  That is going to make they check look bad?  Although it was updating parity as it moved stuff in.

No.   Parity is updated real time so you should never get any parity check errors if everything is good.  Doing such a check is optional and just a confidence building step.

[mattw]

It just came back error free!

 

Thanks guys.