That should work except if there are read errors, that data will then be gone, but most should be recoverable, with some luck all of it.

4 hours ago, JorgeB said:

That should work except if there are read errors, that data will then be gone, but most should be recoverable, with some luck all of it.

How would you handle this once the sync is done? Could the errors be from empty bits? “Bit rot” or likely in the files? Would u copy a little at a time and test? I don’t want too corrupt my array again 

You basically have two options:

• mount the old disk with UD and copy what you can, any files that occupy one or more bad sectors will fail top copy, you will need to skip those files
• option two clone the disk with ddrescue, it will skip the errors and copy the rest from the affected files (in the end it's possible to get a list of possible corrupt files), if they are movies files for example a single bad sector may result in a little glitch playing back the file, that still might be acceptable if there are no backups.

2 minutes ago, JorgeB said:

You basically have two options:

• mount the old disk with UD and copy what you can, any files that occupy one or more bad sectors will fail top copy, you will need to skip those files
• option two clone the disk with ddrescue, it will skip the errors and copy the rest from the affected files (in the end it's possible to get a list of possible corrupt files), if they are movies files for example a single bad sector may result in a little glitch playing back the file, that still might be acceptable if there are no backups.

Will it skip them or will I have to tell it to skip them using UD? So a bulk copy should be fine? Also is a party sync after copying with UD needed right away 

7 minutes ago, Lassley7 said:

Will it skip them or will I have to tell it to skip them using UD?

It will depend on the app you use to copy them, some will stop and ask to retry or skip, some retry a few times and then skip.

3 minutes ago, JorgeB said:

It will depend on the app you use to copy them, some will stop and ask to retry or skip, some retry a few times and then skip.

What app would you recommend? I’ve never had too copy over “bad” data before anytime I’ve had a drive fail I’ve been able too use the proper way of emulating and rebuilding its all replaceable data on that drive so I’d rather skip ANY that could be bad and not risk it “trying multiple times first”

Edited September 1, 20232 yr by Lassley7

I would use rsync, it will auto skip any error after a retry and it's easy to resume if needed.

13 minutes ago, JorgeB said:

I would use rsync, it will auto skip any error after a retry and it's easy to resume if needed.

Will krusader do this? Do you have a recommended tutorial for rsync for my situation? Thanks for all the help recently 

Don't know, never used it.

So I got everything back up and running. I’m going to assume that corrupted data can throw errors again? I moved all the data without issues too a new drive, once the “bad” drive was empty I ran 2 extended smart test and both said passed 0 errors.. doesn’t this likely mean I likely moved bad data back too my array? And the drive was actually never bad… should I run a new parity sync?

poseidon-smart-20230905-0505.zip

Edited September 5, 20232 yr by Lassley7

If there were no errors during the copy I would assume all data is OK.

Just now, JorgeB said:

If there were no errors during the copy I would assume all data is OK.

But IF it’s not, will that corrupt parity again? Folks on discord said unraid doesn’t care about the data only the sectors? Also would you trust putting this drive that was throwing errors back in? Now it reads as a good drive so that doesn’t make sense to me. That tells me the data that was bad on that drive is now back in the array..

7 minutes ago, Lassley7 said:

But IF it’s not, will that corrupt parity again?

Not sure what you mean.

 

Data doesn't corrupt parity, bad disk might, if you do a correcting check.

1 minute ago, JorgeB said:

Not sure what you mean.

 

Data doesn't corrupt parity, bad disk might, if you do a correcting check.

From the beginning we narrowed this issue down to a bad connection/drive. After doing all the steps to recover my data off that drive, I then ran 2 extended smart tests. And they both passed without error. Does this mean the drive was good? And the data that I copied back over to the array was not? I’m confused on how it was a bad drive and now it’s showing good. Should I put this drive back in the array? I posted the smart test above 

42 minutes ago, Lassley7 said:

And they both passed without error. Does this mean the drive was good?

If the disk passed an extended SMART test it should be OK, at least for now.

 

42 minutes ago, Lassley7 said:

And the data that I copied back over to the array was not?

Why would the data not be OK if the disk is good?

 

 

15 minutes ago, JorgeB said:

If the disk passed an extended SMART test it should be OK, at least for now.

 

Why would the data not be OK if the disk is good?

 

 

The disk wasn’t good, that’s why I got into this mess to begin with. That’s why I’m confused as to why it’s now showing good, when nothing has changed other then the data is on a new disk in the array and this one is now empty.

14 minutes ago, Lassley7 said:

The disk wasn’t good, that’s why I got into this mess to begin with.

If the extended SMART test passed the disk is good for now, and if the data copied without any errors there's nothing to make you think it wasn't good.

3 minutes ago, JorgeB said:

If the extended SMART test passed the disk is good for now, and if the data copied without any errors there's nothing to make you think it wasn't good.

Then what was throwing the errors to begin with? It had to be bad disk/data…

Data doesn't throw errors, could have been an intermittent disk issue or a power/connection problem.